Survey analysis system and method

[0002] The present application is related to U.S. patent application Ser. No. 08/937,025, filed Sep. 23, 1997, entitled “Method and Apparatus for Warning a User of Potential Limitations of a Database Request and/or the Results Provided Thereby”; U.S. patent application Ser. No. 08/937,354, filed Sep. 23, 1997, entitled “Method and Apparatus for Using Prior Results When Processing Successive Database Requests”; U.S. patent application Ser. No. 08/937,351, filed Sep. 23, 1997, entitled “Method and Apparatus for Validating a Survey Database”; U.S. patent application Ser. No. 08/937,352, filed Sep. 23, 1997, entitled “Method and Apparatus for Identifying the Coverage of a Test Sequence in a Rules-Based Expert System” now U.S. Pat. No. 6,125,359; and U.S. patent application Ser. No. 08/937,353, filed Sep. 23, 1997, entitled “Method and Apparatus for Detecting an Endless Loop in a Rules-Based Expert System”, all of which are assigned to the assignee of the present invention and incorporated herein by reference.

[0003] 1. Field of the Invention

[0004] The present invention generally relates to data processing systems and more particularly relates to such systems that analyze survey data.

[0005] 2. Description of the Prior Art

[0006] Many businesses have adopted the concept of Total Quality Management (TQM) to help improve profitability, ensure sustained customer loyalty, improve quality of products, etc. Originally developed by E. Deming and others, TQM is now widely accepted in many business circles throughout the world. A cornerstone of most total quality management (TQM) systems is the periodic measurement of identified parameters that relate to various aspects of the business. Through these measurements, managers can identify the areas within the organization that need improvement. Because the measurements are made on a “periodic” basis, managers can gauge the effects of various changes made within the organization over time. The goal of most TQM systems is to make continuous improvement within the organization.

[0007] Businesses often generate and manage an array of measurements for purposes of analyzing the performance of individual operations in relationship to productivity, resource utilization, profitability, etc. One of the most important measurements in many TQM systems is that of customer satisfaction. To measure customer satisfaction, periodic customer feedback is required, usually obtained through customer surveys.

[0008] In many surveys, customers are asked a number of survey questions that are related to various products and/or services provided by a business. For many questions, the customers can respond both with a satisfaction rating and an importance rating. By analyzing the answers to these questions, managers can obtain insights into many areas, including customer loyalty, overall satisfaction, potential weakness, etc.

[0009] It is not uncommon for businesses to conduct their own customer surveys. However, it is increasingly popular to outsource this function to outside vendors who specialize in conducting customer surveys. Today, there are a number of vendors who conduct customer surveys, and provide a survey database to the subscribing business. One such company is Prognostics Inc., located in Menlo Park, Calif. An advantage of using such a vendor is that customer survey data for other companies, including competing companies, can often be obtained. This enables a business to perform comparisons between itself and its competitors.

[0010] Some of the vendors of customer surveys provide the survey data in electronic form, and often provide a software program for accessing the survey database. The software programs facilitate the generation of reports or the like, which can then be used by management. Prognostics is one such vendor. Prognostics provides survey data in a proprietary electronic format, and provides a software program called the Research Analysis Program (RAP) which can access the survey data. RAP can read the survey database, perform data requests, and provide a number of reports.

[0011] A difficulty with many of the software programs provided by customer survey vendors is that only certain data requests can be performed on the survey database. The performance of alternative data requests, which may be more desirable for a particular user, often require the user to execute several of the data requests offered by the vendor's software program, and independently calculate the desired result using the data provided in the vendor's reports. This may be particularly problematic because there are typically only a few “experts” within a business organization that have the knowledge necessary to select the appropriate data requests offered by the vendor's software program, and perform the necessary calculations to produce the desired result. Thus, performing an analysis of a survey database provided by a survey vendor can require a significant amount of time of key personnel to arrive at the desired results.

[0012] This problem is magnified because the concept of Total Quality Management (TQM) requires that “continuous” measurement of selected parameters be performed. In practice, survey data is often taken once-per-week, once-per-month, once-per-year, etc., depending on the particular parameters to be measured. Customer satisfaction, for example, is often measured once-per-month or once-per-year, and requires a new customer satisfaction survey for each measurement. Every time a new customer satisfaction survey is performed, a significant amount of time of key personnel within an organization is often required to re-evaluate the survey database.

[0013] The present invention overcomes many of the disadvantages of the prior art by providing a method and apparatus for analyzing a survey database, wherein a user can perform a wider variety of requests to the survey database without having to independently calculate a desired result. Moreover, additional request types identified by “experts” within an organization are relatively easy to add to the system, and the parameters used by many of the requests can be set by the user at run time adding additional flexibility to the system.

[0014] In accordance with the present invention, these and other advantages are accomplished by using a rules-based expert system for forming and executing requests to a survey database. In a rules-based expert system, a number of rules are provided wherein the rules contain much of the “knowledge” of the experts, and the insights of “decision makers”. This allows “non-expert” users to perform “expert” analysis of client satisfaction data in an accurate and repeatable fashion. This may make data analysis of a survey database faster, more reliable, and cheaper than in the prior art. Moreover, additional request types can be added to the system by simply adding additional rules, and linking the additional rules into the system.

[0015] Preferably, the rules-based expert system uses a number of predefined rules to process a request. Each rule is similar to a conditional statement in a conventional computer program (such as C++). However, the rules-based expert system is equipped with an inference engine, which is a special program for managing rules and applying them as appropriate. By contrast, a conventional program must indicate explicitly when a given conditional statement should be applied.

[0016] It is contemplated that the rules may have variables associated with them. By simply changing the values of the variables at run time, the analysis can be changed or optimized “on the fly”. The underlying code may remain unchanged, and thus may not require recompilation, extensive debugging and/or quality assurance testing. In addition, the rules themselves may be added or changed, rather than just a variable associated with a rule, thereby changing the function of the rules-based expert system. Because the inference engine applies the rules as appropriate rather than in a predetermined order as in conventional code, the basic “algorithm” need not be modified when rules are added or modified.

[0017] In one embodiment of the present invention, a system is provided that has an interface module for accepting a request from a user, and a knowledge module that communicates with the interface module and is capable of accessing the survey database. The knowledge module processes the request at least in part by executing selected ones of a number of predefined rules. An inference engine selects which of the number of predefined rules are actually applied for a particular request.

[0018] The knowledge module may receive a request from the interface module, and determine which of the data elements in the survey database are required to derive the desired response to the user request. This may be accomplished, at least in part, by executing a number of predefined rules in the rules-based expert system, resulting in a number of selected data elements. The selected data elements may then be read from the survey database, and the knowledge module may determine the desired response to the user request at least in part by executing a number of predefined rules in the rules-based expert system. In most cases, the data elements within the survey database correspond to answers to questions in one or more customer surveys.

[0019] To determine the appropriate data elements in the survey database, the knowledge module may first determine the type of analysis requested by the user, and determine which of the data elements in the survey database are required to derive a response to the identified analysis type. Illustrative analysis types include a strength analysis type, a weakness analysis type, a threat analysis type, and an opportunity analysis type.

[0020] The present invention also contemplates providing caveats in response to certain requests. A first illustrative type of caveat is a request caveat. A request caveat may indicate if the request is a valid request. For example, the “experts” of a business organization may determine that a comparison of customer satisfaction between a mainframe computer system and a personal computer system does not provide any useful information. In this instance, this request may be deemed invalid. Preferably the system of the present invention still may process the request, but may issue a request caveat warning the user of any potential issues. It is also contemplated that a request caveat may indicate if any syntax errors have been found.

[0021] A second illustrative type of caveat is a results caveat. A results caveat indicates if the user should view the results with caution. In one example, the knowledge module may determine if the survey database has an insufficient number of selected data elements to derive a statistically significant response to the user request. That is, if there are insufficient data elements in the survey database to produce a statistically significant result, a results caveat may be provided indicating that the results should be viewed with caution. It is noted that the results may still be provided to the user. The results caveats may also identify a group of selected data elements that caused the result to be statistically insignificant. Because the rules in a rules-based expert system include knowledge provided by the “experts”, the request caveats may be provided by executing a number of user request rules provided in a rules-based expert system. The rules may not only determine if a request is valid, but may also indicate to the user the reasons why the user request was determined to be invalid.

[0022] Rather than submitting data requests directly to the survey database, it is contemplated that the knowledge module may assemble a number of data requests, and submit the data requests to a data module. Depending on the user request, different data requests may be required. For example, a request specifying an opportunity analysis type may require a strength analysis for a selected product and a weakness analysis for a competitor's similar product. Each of these analysis types may require different data requests to the survey database. The knowledge module assembles the appropriate data requests, based in the request provided by the user, and provides the data requests to the data module.

[0023] The data module then provides the data request to a report generator. The report generator may be implemented in hardware or as a survey analysis program. The survey analysis program may be provided by the vendor of the survey data, or some other system or program that can access the survey database. This may be desirable when the survey data is in a proprietary format, thus requiring the use of the vendor's survey analysis program.

[0024] In an illustrative embodiment, the report generator accesses the survey database and provides a number of intermediate reports to the rules-based expert system. The intermediate reports typically include a number of data objects therein. The rules-based expert system then uses the requested intermediate reports to calculate and/or assemble one or more reports that are provided to the user.

[0025] Preferably, the survey database includes a number of separate surveys. Each of the number of separate surveys may include data elements that correspond to a number of common questions, and a number of data elements that correspond to a number of questions that are not common to all of the separate surveys. Despite this lack of correlation between some of the separate survey files, it may be desirable to provide data requests and produce results that are based on two or more of the separate surveys. For example, one survey file may include survey data for one year while another survey file may include survey data for another year, and a comparison of the customer satisfaction between the two years may be desirable. Likewise, some of the questions in a given survey may be ordered differently or may be worded differently from another survey.

[0026] To overcome this and other difficulties, the present invention contemplates providing a correlation table which correlates the various questions in the separate survey files. This is preferably accomplished by correlating each question in each survey file to a master listing of generic questions or topics. Alternatively, each question may be correlated to all other similar questions in all other survey files.

[0027] Preferably, the interface module displays a listing of the generic questions, which may be selected by the user when forming a user request. To determine the appropriate generic questions for a selected survey, the interface module may access the correlation table, preferably via the data module, and retrieve the generic questions that correspond to the actual questions in the selected survey. Once a user request is formed, the knowledge module may identify the actual question numbers and the appropriate surveys that correspond to the generic questions provided in the user request. The data requests assembled by the knowledge module preferably specify the actual questions numbers, and the corresponding survey names. Thus, the data requests provided by the knowledge modules preferably identify selected data elements within the survey database for analysis.

[0028] In some cases, it is desirable to provide ordered processing within a rules-based expert system. This can be difficult when using some conventional expert type systems. To overcome this limitation, the present invention contemplates identifying a first set of rules and a second set of rules that are required for performing a task, wherein the first set of rules are to be processed before the second set of rules. The first set of rules are submitted for processing, preferably by an inference engine of the rules-based expert system, and a first result is provided. Thereafter, the second set of rules are submitted to the inference engine. Thus, a method is provided for producing ordered processing within a rules-based expert system.

[0029] Preferably, when the first set of rules are processed by the inference engine, a number of first agenda objects are generated within an object oriented database. Each of the agenda objects may specify a data request and a number of rules. The system then executes each of these agenda objects in a predetermined order. For example, the system first processes a first agenda object. This includes executing the data request associated with the first agenda object, and obtaining the appropriate data elements from the survey database. Thereafter, the number of rules associated with the first agenda object are executed using the inference engine. This provides a number of first results. The subsequent agenda objects are processed in a similar manner. Using this approach, ordered processing can be achieved in a rules-based expert system.

[0030] Finally, it is contemplated that a scripting feature may be provided. As indicated above, Total Quality Management (TQM) often requires periodic evaluation of survey data. Because it may be desirable to make selected requests to a number of survey databases, including subsequent survey databases obtained at a later time, it may be desirable to save a number of requests to a script file. The requests in the script file can then be executed automatically by simply executing the script when subsequent survey databases are received.

[0031] In many cases, the same basic requests are executed on each survey database received. However, selected parameters or system settings may change over time. For example, the definition of a strength may be defined differently over time, possibly requiring a modification to a parameter or system setting. As indicated above, modifying selected parameters may be automatically reflected in the rules that reference the parameter.

[0032] To facilitate the execution of requests stored in a script file using alternative parameter or system settings, the present invention contemplates saving selected user requests independent of the system settings and/or parameters in a user request script. The user request script can then be read, and processed using new parameter or system settings that are currently active at the time of execution. The term “system settings” may include adding, modifying or deleting rules from the rules-based expert system.

[0033] Other objects of the present invention and many of the attendant advantages of the present invention will be readily appreciated as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings, in which like reference numerals designate like parts throughout the figures thereof and wherein:

[0034] FIG. 1 is a block diagram of the computer-based environment of the present invention;

[0035] FIG. 2 is a block diagram showing a first illustrative embodiment of the present invention;

[0036] FIG. 3 is a block diagram showing an illustrative data processing system in accordance the present invention;

[0037] FIG. 4 is a block diagram showing in more detail the execution and control module 58 of FIG. 3;

[0038] FIG. 5 is a block diagram showing in more detail the interface module 52 of FIG. 3;

[0039] FIG. 6 is a block diagram showing a second illustrative embodiment of the present invention;

[0040] FIG. 7 is a block diagram showing an illustrative data processing system in accordance with another embodiment of the present invention;

[0041] FIG. 8 is a diagram showing a preferred analysis control window in accordance with the present invention;

[0042] FIG. 9 is a diagram showing the preferred analysis control window of FIG. 8, with the type of analysis window expanded;

[0043] FIG. 10 is a diagram showing a preferred transaction viewer window, which is displaying the results of a strength type analysis;

[0044] FIG. 11 is a diagram showing the transaction viewer window of FIG. 10, with the Result Caveats button selected;

[0045] FIG. 12 is a diagram showing a preferred explanation viewer window;

[0046] FIG. 13 is a diagram showing the preferred transaction viewer window of FIG. 10, after a subsequent weakness request has been executed;

[0047] FIG. 14 is a diagram showing the preferred analysis control window of FIG. 8, with the survey question selected to be Product Vender, and a response of Unisys;

[0048] FIG. 15 is a diagram showing the resulting transaction viewer window after the execute request button of FIG. 14 is depressed;

[0049] FIG. 16 is a diagram showing the preferred analysis control window with an opportunity analysis type selected;

[0050] FIG. 17 is a diagram showing a preferred SWOT editor window for editing SWOT parameters;

[0051] FIG. 18 is a diagram showing a preferred script editor window in accordance with the present invention;

[0052] FIG. 19 is a table showing an illustrative correlation table;

[0053] FIG. 20 is a functional diagram showing the interaction of the interface module, knowledge module and data module of the preferred embodiment of the present invention;

[0054] FIG. 21 is a flow diagram showing the general execution of a request in accordance with the preferred embodiment of the present invention;

[0055] FIG. 22 is a low level object model of the preferred interface module;

[0056] FIG. 23 is a composite object model for the preferred knowledge module;

[0057] FIG. 24 is a diagram showing a functional model for a SWOT analysis;

[0058] FIG. 25 is a diagram showing a functional model of a strength analysis;

[0059] FIG. 26 is a diagram showing a functional model of a weakness analysis;

[0060] FIG. 27 is a diagram showing a functional model of an opportunity analysis;

[0061] FIG. 28 is a diagram showing a functional model of a threat analysis;

[0062] FIG. 29 is a diagram showing a functional model of a survey validation analysis;

[0063] FIG. 30 is a diagram showing a functional model of the preferred data module;

[0064] FIG. 31 is a diagram showing a consolidated low level object model of the preferred data module;

[0065] FIG. 32 is a diagram showing a functional model of a preferred correlation table manager;

[0066] FIG. 33 is a table showing the execution of an illustrative direct RAP request, with a problem detected in the request;

[0067] FIGS. 34A-34B show the execution of an illustrative direct RAP request, when no problems are detected in the request or when the user indicates that the request should be executed despite the detected problem;

[0068] FIGS. 35A-35B show the execution of an illustrative strength or weakness analysis, when problems are detected in the request;

[0069] FIGS. 36A-36C show the execution of an illustrative strength or weakness analysis, when no problems are detected in the request or when the user indicates that the request should be executed despite the detected problem;

[0070] FIGS. 37A-37C show the execution of an illustrative threat or opportunity analysis, when no problems are detected in the request or when the user indicates that the request should be executed despite the detected problem;

[0071] FIG. 38 is a table showing the execution of an illustrative survey validation request, with request problems detected;

[0072] FIGS. 39A-39B show the execution of an illustrative survey validation request, when no request problems are detected or when the user indicates that the request should be executed despite the detected problems;

[0073] FIG. 40 is a table showing the execution of an illustrative SWOT parameter update using the SWOT editor of FIG. 17;

[0074] FIG. 41 is a flow diagram showing an illustrative method of the present invention;

[0075] FIG. 42 is a flow diagram showing another illustrative method of the present invention;

[0076] FIG. 43 is a flow diagram showing the method of FIG. 42, with elements 1206, 1208 and 1210 of FIG. 42 replaced with elements 1220 and 1222 of FIG. 43;

[0077] FIG. 44 is a flow diagram showing another variation of the method shown in FIG. 42, with element 1206 of FIG. 42 replaced with element 1228 of FIG. 44;

[0078] FIG. 45 is a flow diagram showing an illustrative method for providing ordered processing in a rules based expert system;

[0079] FIG. 46 is a flow diagram showing another illustrative method for providing ordered processing in a rules based expert system;

[0080] FIG. 47 is a flow diagram showing an illustrative method for providing scripting of requests;

[0081] FIG. 48 is a flow diagram showing an illustrative method for increasing the performance of a survey database analysis system by using previously generated results;

[0082] FIG. 49 is a flow diagram showing an illustrative method for identifying rules that were executed in a rules based expert system;

[0083] FIG. 50 is a flow diagram showing an illustrative method for detecting when a rules based expert system is stuck in a loop;

[0084] FIG. 51 is a table showing illustrative analysis types, and the corresponding rule sets that are associated therewith;

[0085] FIGS. 52A-52C show a table of preferred rule sets, and the corresponding rules that are associated therewith;

[0086] FIGS. 53A-530 show a table of preferred rule names along with corresponding rule comments;

[0087] FIGS. 54A-54F shows an illustrative listing of the object oriented database before any analysis runs are executed; and FIGS. 55A-55G shows an illustrative listing of the object oriented database before after a strength analysis type is executed.

[0088] The detailed descriptions which follow are presented largely in terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art.

[0089] An algorithm is here, generally, conceived to be a self-consistent sequence of steps leading to a desired result. These steps are those requiring physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical or magnetic signals-capable of being stored, transferred, combined, compared, and otherwise manipulated. It proves convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers or the like. It should be kept in mind, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities.

[0090] Furthermore, the manipulations performed are often referred to in terms, such as adding or comparing, which are commonly associated with mental operations performed by a human operator. No such capability of a human operator is necessary, or desirable in most cases, in any of the operations described herein which form part of the present invention; the operations are machine operations. Useful machines for performing the operations of the present invention include general purpose digital computers or other similar devices. In all cases, it should be kept in mind the distinction between the method of operations in operating a computer and the method of computation itself. The present invention relates to method steps for operating a computer in processing electrical or other (e.g., mechanical, chemical) physical signals to generate other desired physical signals.

[0091] The present invention also relates to apparatus for performing these operations. This apparatus may be specially constructed for the required purposes or it may comprise a general purpose computer as selectively activated or reconfigured by a computer program stored in the computer. The algorithms presented herein are not inherently related to a particular computer system or other apparatus. In particular, various general purpose computer systems may be used with computer programs written in accordance with the teachings of the present invention, or it may prove more convenient to construct more specialized apparatus, to perform the required method steps. The required structure for such machines will be apparent from the description given below.

[0092] In sum, the present invention preferably is implemented for practice by a computer, e.g., a source code expression of the present invention is input to the computer to control operations therein. It is contemplated that a number of source code expressions, in one of many computer languages, could be utilized to implement the present invention. A variety of computer systems can be used to practice the present invention, including, for example, a personal computer, an engineering work station, an enterprise server, etc. The present invention, however, is not limited to practice on any one particular computer system, and the selection of a particular computer system (including a computer network) can be made for many reasons.

[0093] FIG. 1 is a block diagram of the computer-based environment of the present invention. A user 10 interacts, with the data processing system 12, which includes a processor 14, which executes operating system software, as well as application programs including the present invention. The processor is found in all general purpose computers and almost all special purpose computers. The data processing system 12 is intended to be representative of a category of data processors suitable for supporting survey analysis operations. In the preferred embodiment, the data processing system 12 is an IBM compatible personal computer (PC) running Windows 3.1, 95 or NT.

[0094] The user 10 enters information into the data processing system by using a well-known input device 16, such as a mouse, keyboard, or a combination of the two devices. It should be understood, however, that the input device may actually consist of a card reader, magnetic or paper tape reader, or other well-known input devices (including another computer system). A mouse or other cursor control device is typically used as an input device as a convenient means to input information to the data processing system to select command modes, edit input data, and the like. Visual feedback is given to the user by showing characters, graphical symbols, windows, buttons or the like on display 18. The display is used to display messages and symbols to the user. Such a display 18 may take the form of any of several well-known varieties of CRT displays. Display 18 may also be in the form of a printer, file storage system, magnetic tape, etc. The software executed by the processor 14 stores information relating to the operation thereof in memory 20. The memory may take the form of a semiconductor memory, magnetic disks, optical disks, magnetic tape or other storage device.

[0095] FIG. 2 is a block diagram showing a first illustrative embodiment of the present invention. The block diagram is generally shown at 30, and includes an expert system 32, which communicates with a survey database 42. The expert system 32 includes an interface module 34, a knowledge module 36, and a data module 38. A user 40 communicates with the system via interface module 34, as shown. The interface module 34 helps the user 40 assemble a user request, which is provided to knowledge module 36.

[0096] Knowledge module 36 checks the user request both for syntax and content. This is preferably done using an inference engine (not shown), which executes a number of rules. The rules contain “knowledge” provided by a number of survey database analysis experts. Knowledge module 36 then stores a number of request caveats for later reference. The request caveats indicate if the user request provided by the interface module is a proper request based on the knowledge within the rules. Knowledge module 36 then assembles an appropriate survey request, and provides the request to data module 38.

[0097] Data module 38 accesses the survey database 42 and provides the requested data objects back to knowledge module 36. Knowledge module 36 processes the requested data objects and provides a number of reports back to interface module 34. Knowledge module 36 may also assemble a number of result caveats, and provide the result caveats to interface module 34. Knowledge module 36 preferably processes the requested data objects using an inference engine within a rules-based expert system. The user 40 can view the request caveats, reports and result caveats via interface module 34.

[0098] When a number of survey results are included in survey database 42, it may be necessary to provide a correlation table 44 which correlates the various questions of each of the surveys. Data module 38 may read the correlation table 44, and determine the appropriate requested data objects therefrom. This may occur, for instance, when the request provided by knowledge module 36 requires access to survey data from two separate years. The surveys used during these two separate years may have different questions or the questions may be in a different order relative to one another. In this instance, correlation table 44 may provide a correlation between the survey questions in the various surveys.

[0099] In addition, it is contemplated that the interface module 34 may display a listing of the generic questions, which may be selected by the user when forming the user request. To determine the appropriate generic questions for a selected survey, the interface module may access the correlation table 44 via the data module, and retrieve the generic questions that correspond to the actual questions in the selected survey.

[0100] Once a user request is formed, the knowledge module 36 may access the correlation table 44 to identify the actual question numbers and the appropriate surveys that correspond to the generic questions provided in the user request. The data requests provided by the knowledge module 36 to the data module 38 preferably specify the actual questions numbers, and the corresponding survey names in the survey database 42 to access. The data requests provided by the knowledge module 36 thus may identify selected data elements within the survey database 42 for analysis.

[0101] The data module 38 accesses the survey database 42 and provides the resulting data objects to knowledge module 36. Knowledge module 36 determines the requested response and provides the results to interface module 34.

[0102] FIG. 3 is a block diagram showing an illustrative data processing system in accordance with the present invention. The data processing system includes an interface module 52, a knowledge module 56, and a file storage area 83. The interface module 52 communicates with a user via interface 54. The interface module 52 may aid the user in assembling a survey database request.

[0103] Interface module 52 provides the resulting request to execution and control module 58 of knowledge module 56. The execution and control module 58 may check the requests received from interface module 52 by initiating execution of a number of rules 72 via inference engine 60. That is, execution and control module 58 may initiate inference engine 60 via interface 78, wherein inference engine 60 may execute a number of rules to check the request. Execution and control module 58 may then receive the results from inference engine 60 via interface 80, wherein execution and control module 58 may assemble a number of request caveats. Execution and control module 58 provides these request caveats to request caveat generator 64 via interface 82. Request caveat generator 64 stores the request caveats in the file storage area 83, as shown at 86.

[0104] Execution and control module 58 may also read survey database 68 via interface 70. After the appropriate data elements are read from the survey database 68, execution and control module 58 may again direct inference engine 60 to execute a number of rules 72. The rules may generate the desired results, and may further check the validity of the results. Execution and control module 58 may then provide the results to report generator 62 via interface 82, and any result caveats to result caveat generator 66. Report generator 62 and result caveat generator 66 may store the reports 84 and result caveats 88 in the file storage area 83, respectively.

[0105] As indicated above, when a request from interface module 52 requires data from more than one survey, it may be desirable to correlate various questions of the various surveys. To do so, execution and control module 58 may provide a request to correlation module 90 via interface 91. Correlation module 90 may read correlation tables 92 via interface 94, and provide the desired correlation data between the various survey questions.

[0106] In addition, the interface module 52 preferably displays a listing of the generic questions for selection by the user when forming the user request. To determine the appropriate generic questions for a selected survey, the interface module 52 may access the correlation module 90 via execution and control module 58, and retrieve the generic questions that correspond to the actual questions in the selected survey.

[0107] Once a user request is formed, the knowledge module may identify the actual question numbers and the appropriate surveys that correspond to the generic questions provided in the user request. The data requests provided by the knowledge module, either directly or indirectly to the survey database, preferably specify the actual questions numbers and the corresponding surveys to access. The data requests provided by the knowledge module may thus identify selected data elements within the survey database for analysis.

[0108] FIG. 4 is a block diagram showing in more detail the execution and control module 58 of FIG. 3. The execution and control module 58 may include an analysis type module 122, a data element determining module 124, and a response determining module 128. The analysis type module 122 determines the analysis type of the user request provided by the interface module. In a preferred embodiment, the analysis types include a strength analysis type, a weakness analysis type, an opportunity analysis type, a threat analysis type, etc.

[0109] Once the analysis type is determined, data element determining module 124 determines which of the data elements in the survey database are to be accessed. For some requests, the request itself may identify the data elements to be accessed. That is, the user request may indicate which questions from which surveys are to be processed. For other requests, however, the correlation table may be required to identify the appropriate questions within each of the surveys.

[0110] Data element determining module 124 provides a corresponding request to data module 126, as shown. Data module 126 accesses the survey database and provides the resulting data objects to response determining module 128. Response determining module 128 determines the requested response and provides the results to the report and caveat generators (see FIG. 3).

[0111] The desired response can be dependent upon a number of analysis parameters 130. In accordance with one feature of the present invention, the analysis parameters 130 are user programmable. It is contemplated that response determining module 128 may determine the appropriate response by executing a number of rules via inference engine 60. The analysis parameters 130 preferably are referenced by a number of these rules. Thus, the analysis parameters 130 may affect the results provided by the rules, and therefore, the response provided by response determining module 128.

[0112] For each request, data module 126 may store the results in prior results file 134. A search block 132 may be used to search the prior results file 134 before accessing the survey database. If a request from data element determining module 124 requires access to results that were already generated for a previous request, data module 126 may read the appropriate prior results file 134 and provide the appropriate results directly to response determining module 128. Because accessing the survey database may require more time than simply accessing the prior results file 134, this feature can save considerable time in processing a request. This is particularly true when data module 126 accesses the survey database via a survey analysis program, which must calculate and assemble a number of intermediate reports which are provided to response determining module 128.

[0113] FIG. 5 is a block diagram showing in more detail the interface module 52 of FIG. 3. Interface module 52 may include a request module 140, and a response module 142. The request module may aid the user in assembling the user requests which are provided to knowledge module 56 via interface 76. The response module 142 aids the user in reading and viewing the reports and various caveats received via interface 98.

[0114] It is contemplated that interface module 52 may include a script module 144. Script module 144 is preferably coupled to request module 140, as shown. In a scripting mode, request module 140 may provide the assembled user request to script module 144, rather than to knowledge module 56. Script module 144 may store a number of assembled user requests in a number of script files 146, as shown. Upon a user's command, the assembled requests within a script file 146 may then be provided to knowledge module 56 from request module 140 via interface 76.

[0115] Accordingly, this feature may allow a user to build a number of script files 146 without having to wait for the knowledge module to process the results. Thereafter, the user may execute all of the assembled requests within a script by simply executing one of the script files 146. The request module 140 may then sequentially provide the assembled user requests within a selected script file to the knowledge module 56 in a batch like mode.

[0116] FIG. 6 is a block diagram showing a second illustrative embodiment of the present invention. In this embodiment, a rules based expert system is utilized. Rules are similar to conditional statements in a conventional computer program. However, rules are typically executed using an inference engine, which manages the application of the rules. A conventional program, by contrast, has to indicate explicitly when given conditional statements should be applied. The rules are defined by the application developer and are incorporated into the rules-based expert system.

[0117] The preferred rules based expert system is the KAPPA-PC programming environment. KAPPA-PC is available from IntelliCorp, Inc., and provides an object-oriented programming environment with an inference engine for rule-based reasoning. A further discussion of the KAPPA-PC programming environment can be found in the KAPPA-PC on-line help files, which are incorporated herein by reference.

[0118] The interface module 168, knowledge-module 170 and data module 196 are all preferably implemented using the KAPPA-PC programming language KAL. The interface module 168 helps a user 174 assemble a desired survey request, and preferably includes all necessary objects and methods to interact with the user including all input screens, RAP command prompts, analysis processing values (object attributes), response windows, etc.

[0119] To help the user focus the requests, the interface module 168 may allow the user to select one of a number of surveys, and any number of certain questions from the selected survey. Because corresponding questions in selected surveys may differ from one survey to another, the interface module 168 preferably displays a listing of generic questions for selection by the user when forming the user request. To determine the appropriate generic questions for a selected survey, the interface module 168 may access the correlation table 184 via data module 172, and retrieve the generic questions that correspond to the actual questions in the selected survey.

[0120] After a request is assembled, the interface module 168 provides the user requests to knowledge module 170. Knowledge module 170 determines which RAP reports will be required to derive the desired result. Knowledge module 170 may also determine a number of request caveats, as described above. After determining which RAP reports are required to derive the desired result, the knowledge module 170 provides appropriate data requests to data module 172 via interface 182. Data module 172 provides the data requests to survey analysis module 164 via interface 188. Survey analysis module 164 executes the data requests. This typically includes accessing the survey database 166 and providing the requested RAP reports back to data module 172 via interface 190.

[0121] Data module 172 may then parse the RAP reports received from the survey analysis module 164, and provide a number of data objects to knowledge module 170. Knowledge module 170, using a number of rules, may then determine the requested results, and provide the requested results to interface module 168. Knowledge module 170 may also provide a number of result caveats to interface module 168 via interface 194. The user 174 may then view the request caveats, the desired results, and the result caveats.

[0122] In a preferred embodiment, the survey analysis module is implemented using the Research Analysis Program (RAP) available from Prognostics, Inc., located in Menlo Park, Calif. RAP is a software program developed by Prognostics to access a survey database and provide a number of reports. These reports include a correlation report, a distribution report, a GAP report, a multi-GAP report, etc. The RAP manual, available from Prognostics, Inc., describes these various reports, and is incorporated herein by reference.

[0123] Generally, the correlation report determines the correlation between the importance and/or satisfaction in one segment of the survey database with the importance and/or satisfaction in another. The distribution report provides the number of responses in a segment of the survey database. The GAP report identifies the gaps between the importance and satisfaction responses for a particular segment of the survey database. The multi-GAP report identifies the gaps between the importance and satisfaction responses for a range of questions in the survey database.

[0124] While these RAP reports can be useful in analyzing survey data, it has been found that more complex reports are often desired. To produce these more complex reports, it has been necessary to generate selected reports offered by the RAP program, and then manually calculate the desired data elements therefrom. This is often tedious and error prone, particularly since the desired reports may have to be updated each time a new survey database is received.

[0125] For those situations where the reports provided by the RAP program are sufficient, the interface module 168 may allow a user 174 to provide direct RAP requests to knowledge module 170. Knowledge module 170 may pass the direct RAP requests to data module 172. Data module 172 may then pass the direct RAP requests to the RAP program 164. The RAP program may execute the direct RAP requests, and provide corresponding reports to data module 172. Data module 172 may then provide the reports to the interface module 168, via knowledge module 170. It is contemplated that the knowledge module 170 may generate request caveats and/or result caveats for the direct RAP requests.

[0126] For complex requests, interface module 168 may assemble a user request and provide the user request to knowledge module 170 via interface 180. Knowledge module 170 performs a check of the user request and provides a number of request caveats to interface module 168 via interface 194. Preferably, knowledge module 170 determines the request caveats by executing a number of rules using the inference engine within the KAPPA-PC environment.

[0127] Thereafter, the knowledge module 170 determines the appropriate data elements that are required to respond to the user request by executing a number of rules. The Knowledge Module may include references to a number of rules, some of which are generic (executed for each complex command), and others that are intended to be executed only for certain commands. For this reason, the knowledge module uses the KAPPA-PC notion of a rule-set for rule-based execution.

[0128] In addition, an object called “inference_network” may be provided in the object oriented database (see FIG. 54D and FIG. 55E). The inference_network objects may contain many multi-valued slots (lists), where each slot corresponds to a rule set. The list of rules to be executed in each instance may be stored in the slot name, and the knowledge module may determine which rule sets to invoke.

[0129] As indicated above, the user request provided by the interface module preferably includes the text of a generic question. To identify the actual question number within the selected survey, the knowledge module 170 may access the correlation table 184 via the data module 172. The correlation table may provide a cross-reference between the generic question text and the corresponding actual question numbers in each of the surveys. In this way, the knowledge module 170 may identify the data elements within, the survey database that are required to respond to the user request. A further discussion of the correlation table 184 can be found with reference to FIG. 32.

[0130] The knowledge module 170 provides a number of data requests to data module 172, preferably identifying the appropriate data elements within the survey database. Data module 173 forwards the data requests to the survey analysis module 164, wherein the data requests are executed. The survey analysis module 164 provides a number of resulting reports to data module 172. In the preferred embodiment, the reports provided by the survey analysis module 164 are in an ASCII format. Data module 172 may parse the reports into a number of data objects, and provides the data objects to knowledge module 170.

[0131] Using the data objects, knowledge module 170 determines the requested results. Knowledge module 170 may also determine a number of result caveats. The results and result caveats may then be provided to interface module 168.

[0132] It is contemplated that data module 172 may store the data objects received from survey analysis module 164 in a prior results file 198. Once a request is received by data module 172 from knowledge module 170, data module 172 may search the prior results file 198 via search engine 196 to determine whether the requested data objects have already been generated and stored. If so, data module 172 may provide the requested data objects from prior results file 198 directly to knowledge module 170, rather than regenerating the data objects via survey analysis module 164. This may decrease the execution time for a corresponding request, particularly since the re-use of prior results objects can occur for different request types.

[0133] FIG. 7 is a block diagram showing an illustrative data processing system 210 in accordance with another embodiment of the present invention. The data processing system 210 includes an interface module 212, a knowledge module 216, and a file storage area 246. The interface module 212 communicates with a user via interface 214. The interface module 212 aids the user in assembling survey database requests, as described above. Interface module 212 provides the resulting request to execution and control module 218 of knowledge module 216.

[0134] Preferably, execution and control module 218 checks the requests received from interface module 212 by initiating execution of a number of rules 236 via inference engine 220. That is, execution and control module 218 may provide a request to inference engine 220 via interface 232, wherein inference engine 220 may execute a number of selected rules. Execution and control module 218 may then receive the results from inference engine 220 via interface 234, wherein execution and control module 218 may determine a number of request caveats. Execution and control module 218 provides these request caveats to request caveat generator 224 via interface 244. Request caveat generator 224 stores the request caveats in the file storage area 246, as shown at 250.

[0135] Execution and control module 218 may also provide a corresponding request to data module 228. Data module 228 determines the appropriate requests to submit to survey analysis module 240. Survey analysis module 240 receives the requests from data module 228, and performs the requests on the survey database 242. The survey analysis module receives the results and provides the results to data module 228. Data module 228 forwards the results to execution and control module 218. It is contemplated that data module 228 may process the results prior to forwarding the results to execution and control module 218.

[0136] Execution and control module 218 may initiate execution of the inference engine 220 to check the results. The inference engine 220, by executing a number of the rules 236, may provide a number of result caveats. Execution and control module 218 may then either provide the results directly to report generator 222 via interface 244, or perform further processing on the results and provide those results to report generator 222. Also, execution and control module 218 may provide the result to result caveat generator 226. Report generator 222 and result caveat generator 226 may store reports 248 and result caveats 252 in the file storage area 246.

[0137] As indicated above, when a request from interface module 212 requires data from more than one survey, it may be desirable to correlate various questions of the various surveys. To do so, execution and control module 218 may provide a request to data module 228, which may then access correlation module 254. Correlation module 254 may read correlation tables 256, and provide the desired correlation data between the various survey questions.

[0138] In addition, the interface module 212 preferably displays a listing of the generic questions for selection by the user when forming the user request. To determine the appropriate generic questions for a selected survey, the interface module 212 may access the correlation module 254 via data module 228, and retrieve the generic questions that correspond to the actual questions in the selected survey. Finally, it is contemplated that execution and control module 218 may have a count block 258. Count block 258 may count the number of times each rule or rule set is executed by inference engine 220. If a rule or rule set is executed more than a predetermined number of times, the count block 258 interrupts the inference engine 220. This provides an efficient way to detect when the inference engine 220 is stuck in a loop.

[0139] FIG. 8 is a diagram showing a preferred analysis control window in accordance with the present invention. The control window is generally shown at 300, and includes a number of regions. A first region is the command mode region 302. The command mode region 302 includes a direct RAP request selection and a complex request selection. The direct RAP request selection allows direct RAP requests to be provided to the research analysis program (RAP). When the direct RAP request is selected and the execute request button 310 is depressed, a direct RAP request is executed by RAP, and the corresponding RAP reports are generated.

[0140] In the illustrative diagram, the command mode region 302 shows the complex request mode selected. In this mode, analysis type region 306 provides a number of complex mode analysis types for selection (see, for example, FIG. 9). In the illustrative diagram, a strength type analysis is selected.

[0141] A survey name region 304 is also provided. The survey name region 304 allows a user to select one of a number of surveys from a survey database. In the illustrative diagram, a survey named CS96NA is selected. The survey question region 308 displays the survey questions that correspond to the selected survey. Preferably, the survey question region 308 displays a listing of generic questions that correspond to the actual question in the selected survey. This provides consistency to the questions listed in the survey question region 308, regardless of the selected survey. The user may identify one or more of the items in survey question region 308 to further focus the analysis. None are selected in the illustrative diagram of FIG. 8.

[0142] FIG. 9 is a diagram showing the analysis control window of FIG. 8, with the type of analysis region 306 expanded. As indicated above, a number of analysis types can be selected by the user. In the illustrative diagram, a strength analysis is selected, as shown at 320. The resulting request is shown in the request region 322. Other complex analysis types include weakness, opportunity, threat, behavior, GAP, importance, satisfaction, etc.

[0143] FIG. 10 is a diagram showing a preferred transaction viewer window, which is displaying the results of the strength analysis request of FIG. 9. The execution of each request produces a transaction. The transaction viewer window 330 displays the results of each transaction. In the illustrative diagram, a transaction which has been given the name TR_0 is displayed, as shown at 332. The request information that was used to generate the transaction is shown in the request region 338. The request region 338 indicates that the results shown correspond to a strength type analysis, using a survey named CS96NA, and using the corporate baseline algorithm.

[0144] A view selection region is provided at 335. The results button 334 is shown highlighted, indicating that the results are shown in the response transcript region 336, rather than the request caveats or result caveats.

[0145] FIG. 11 is a diagram showing the transaction viewer window of FIG. 10, with the Result Caveats button selected. When the result caveat button 340 is selected, the response transcript region 342 displays the result caveats for the selected transaction. The request caveat button 341, if selected, would display the request caveats for the selected transaction.

[0146] The transaction viewer window also may include an explain button 344, as shown. The explain button, when selected, displays the rule sets and rules that were activated by the inference engine during the execution of the selected transaction. The display provided by the explain button is further described with reference to FIG. 12.

[0147] FIG. 12 is a diagram showing an explanation viewer window. The explanation viewer window is activated by the explain button of the transaction viewer window (see, FIG. 11). The explanation viewer window has a rule set region 350 that shows the rule sets that were activated during a selected transaction. When a particular rule set is selected within the rule set region 350, such as RS_XQT_STRENGTH_4 as shown at 352, the rules within that rule set are displayed in rules region 354. For example, the rules in rule set RS_XQT_STRENGTH_4 are shown at 354.

[0148] By selecting a particular rule, for example, rule S_SET_REPORT as shown at 356, the corresponding comment for that rule is displayed in the rule comments region 358. Accordingly, the user can determine which rule sets and rules were activated for a selected transaction, and can further view the comments for each of the rules.

[0149] Each rule typically includes an IF clause, THEN clause, and a COMMENT. As indicated above, the comment is accessed through the Explain button by selecting a rule within a rule set. An “Explain” object instance is created for every transaction, and includes the list of rule sets invoked during the transaction.

[0150] FIG. 13 is a diagram showing the transaction viewer window of FIG. 10 after a subsequent weakness request is executed. The transaction region now shows two transactions including TR_0 and TR_1, with TR_1 selected. The corresponding results are shown at 366 in the response transcript region 366. By selecting the appropriate transaction in the transaction region, the corresponding results are displayed in the response transcript region 366.

[0151] FIG. 14 is a diagram showing the analysis control window of FIG. 8, with the Product Vendor question selected. As can be seen, a complex request is selected at 372, and a strength analysis type is selected at 376. The survey name CS96NA is selected at 374, which provides a number of survey question selections in the survey question region. As indicated above, a number of generic questions are displayed in the survey question region.

[0152] In the illustrative diagram, the generic question “product vendor” is selected at 378. The available answers to the product vendor question for the survey CS96NA are displayed in the response region adjacent to the survey question region. The available answers are obtained by accessing a MQA routine. Preferably, the MQA routine accesses a MQA table that is provided by the survey database vendor. The MQA table stores the available answers to each question in a survey.

[0153] The user may select one or more of the product vendor responses. In the illustrative diagram, the UNISYS product vendor is selected at 380. The request region 382 displays the resulting request. The user may execute the request by depressing the Execute Request button 384.

[0154] FIG. 15 is a diagram showing the resulting transaction viewer window after the execute request button of FIG. 14 is depressed. A description of the request is shown at 394, indicating that a product vendor of Unisys is selected. The corresponding results are shown in the response transcript region 398.

[0155] FIG. 16 is a diagram showing an illustrative analysis control window with an opportunity analysis type selected. The complex request command mode is selected at 402, and an opportunity analysis type is selected as shown at 406. An opportunity analysis determines those areas that a particular vendor has an opportunity for gaining market share over other vendors. In the example shown, the primary request selects the survey name CS96NA at 404, and a Product Vendor of UNISYS at 408 and 410, respectively.

[0156] The secondary request selects survey CS96NA as shown at 412, and a Product Vendor of IBM at 414. The resulting opportunity request is shown in the request region 418.

[0157] If a product vendor is selected that is not in competition with, or is otherwise determined to be not compatible with UNISYS, for example, then the system may generate a request caveat indicating the incompatibility. In another example, if an opportunity request is assembled which compares two product types which are incompatible (like a main frame versus a PC), then the system may generate a request caveat indicating the incompatibility. It is contemplated that a user may select a different survey name, survey question and/or survey answer.

[0158] FIG. 17 is a diagram showing a SWOT editor window for editing SWOT parameters. Preferably, some of the rules within the expert system have parameters or variables that can be set or changed by the user. The SWOT editor window 440 provides an interface for changing these variables and/or parameters. The SWOT editor window first allows a user to select one of a number of SWOT algorithms. The acronym “SWOT” refers to one of the Strength, Weakness, Opportunity or Threat analysis types.

[0159] The first SWOT algorithm shown is entitled client significant. The client significant algorithm (importance x satisfaction) analyzes the client satisfaction data using the hypothesis that the relationship between importance and satisfaction reveals whether the vendor is meeting the client's expectations. On the opposite end of the continuum, a client that rates a vendor low on both importance and satisfaction, thus having a low satisfaction and importance product, is indicating to the vendor that the vendor is systematically out of sync with the client's expectations. The result of the client significant algorithm is a continuum of questions ordered on the cumulative effect of the importance/satisfaction product, and allows the user to evaluate the need for changes in business unit strategy.

[0160] The second SWOT algorithm is entitled repurchase contributors. The repurchase contributors algorithm (using MGAP times correlation to repurchase, filtered for high importance, low GAP, and high correlation to purchase intent) provides the vendor with an analysis of the client's importance/satisfaction GAP and the correlation with the client's intent to repurchase the vendor's products. This algorithm uses only paired responses (questions that have both importance and satisfaction responses) to perform the analysis. The hypothesis used by this analysis type is that the most important things to improve are those things that correlate with the client's intent to repurchase.

[0161] The next algorithm selection is entitled client satisfiers. The client satisfiers algorithm (using MGAP times correlation to repurchase filtered for high importance, low GAP and high correlation to satisfaction) provides the vendor with the client's importance/satisfaction GAP in a particular area, and the correlation thereof with the client's overall satisfaction with the vendor. This algorithm also only uses paired responses to perform the analysis. The hypothesis of the client satisfiers analysis type is that the most important things to improve are those things that correlate with the client's overall satisfaction.

[0162] The next algorithm selection is entitled corporate baseline (overall satisfaction). The corporate baseline algorithm (overall satisfaction) uses the STAT GAP times correlation to repurchase filtered for high importance, low GAP, and high correlation to satisfaction. The corporate baseline algorithm (overall satisfaction) provides the vendor with the client's importance/satisfaction GAP, and the correlation thereof with the client's overall satisfaction with UNISYS. This algorithm uses all importance and satisfaction responses in the sample set in order to perform the analysis. The hypothesis of this analysis is that the most important areas to improve are those things that correlate with the client's overall satisfaction.

[0163] The next algorithm selection is entitled corporate baseline (repurchase). The corporate baseline (repurchase) algorithm uses MGAP times correlation to repurchase filtered for high performance, low GAP, and high correlation to repurchase intent. The corporate baseline algorithm (repurchase) provides the vendor with the client's importance/satisfaction GAP, and the correlation thereof with the client's intent to repurchase the vendor's products. This algorithm uses all,importance and satisfaction responses in the sample set in order to perform the analysis. The hypothesis is that the most important areas to improve are those things that correlate with the client's intent to repurchase.

[0164] The last algorithm selection is entitled custom. The custom analysis type allows the vendor to configure the analysis based on various SWOT parameters including the correlation question, correlation cut-off, GAP, importance, etc.

[0165] The SWOT editor window 440 shows the corporate baseline (overall satisfaction) selected at 442. With this selected, a number of corresponding system settings and/or parameters may be changed by the user, including the report cardinality 440, the minimum value answers 446, the strength composite 450, and the weakness composite 448. These parameters may be used by selected rules when a request is executed.

[0166] Each of the algorithms in the SWOT editor-screen is assigned a unique number, X=1 . . . n, which corresponds with rule sets RS init_strength_X, RS_xqt_strength_X, etc. Each algorithm therefore may have an unlimited number of rules associated with initialization and execution in order to generate the desired composite value. The “Custom” or “build your own” algorithm uses the values from the SWOT Editor objects as parameters in the rules associated with that selection.

[0167] FIG. 18 is a diagram showing a script editor window in accordance with the present invention. The script editor window allows the user to manage scripts of requests. Referring briefly back to FIG. 16, the analysis control window allows the user to copy a request to a script via button 489, rather than execute the request. Thus, a number of requests can be assembled by the user and saved to a script file. At a later date, for example when a subsequent survey database arrives, the user may execute one or more of the script files.

[0168] Referring specifically to FIG. 18, a script entity may be selected at 472. Once selected, the contents of the script entity are displayed in the request window 474. The script can be cleared via button 476. The order of the requests within the script entity can be rearranged by deleting or moving a currently selected request via buttons 478, 480, and 482. The resulting script may be saved via button 486. Finally, a particular script may be loaded or appended to a current script via button 484. The script may be executed by simply depressing the execute script button 488.

[0169] FIG. 19 is a table showing an illustrative correlation table. The correlation table is generally shown at 500 and includes a generic heading column 502, and a number of survey columns 504, 506, 508 and 510. Each of the survey columns corresponds to a particular survey within the survey database. For example, the survey shown in column 504 may have been taken in 1995, while the survey shown in column 506 may have been taken in 1996. On one embodiment of the present invention, the correlation table correlates each of the questions within each survey to a generic question. For example, generic heading-1 corresponds to question 122 for surveys S-1 and S-2, and question 143 for surveys S-3 and S-4.

[0170] Under some circumstances, some of the surveys may not have a question that correspond to a generic heading. For example, surveys S-1 and S-2 do not have questions that correspond to generic heading-2. By using such a correlation table, questions within a particular survey can be identified as corresponding to a request that relates to a particular generic heading.

[0171] As indicated above, a result caveat may be provided when a comparison is requested between, for example, surveys S-2 and S-3 with respect to the area addressed by heading-2. That is, the system may provide a result caveat that indicates there is insufficient data in survey S-2 to provide a statistically significant result, if appropriate.

[0172] FIG. 20 is a functional diagram showing the interaction of the interface module, knowledge module and data module of the preferred embodiment of the present invention. The diagram is generally shown at 548. The interface module 552 receives user inputs, user script inputs, server batch requests, etc., as shown.

[0173] To help the user focus the requests, the interface module 552 may allow the user to select one of a number of surveys, and certain questions from the selected survey. Because corresponding questions in selected surveys may differ from one survey to another, the interface module 552 preferably displays a listing of generic questions for selection by the user when forming the user request. To determine the appropriate generic questions for a selected survey, the interface module 552 may access the correlation table 559 via data module 554 and interface 556, and retrieve the generic questions that correspond to the actual questions in the selected survey via interface 557.

[0174] Likewise, the interface module 552 may allow the user to select one of a number of answers to a selected question. The available answers are obtained by accessing a RAP MQA file 563 via data module 554 and interfaces 556 and 557. The RAP MQA file includes the available answers to each question in a survey.

[0175] After a request is assembled, the interface module 552 provides the requests to knowledge module 558 via interface 560. Knowledge module 558 may then determine a number of request caveats for the user request, preferably using a number of rules executed by a rules-based expert system. The request caveats may then be provided back to interface module 552 via interface 572.

[0176] Knowledge module 558 determines which data requests are required to provide the requested results. When determining the appropriate data requests to access the survey database, knowledge module 558 may gain access to the correlation table 559 via data module 554. Correlation table 559 may provide a correlation between generic questions provided in the user request to question numbers in the various surveys. Thus, knowledge module 558 may determine which data elements in the survey database to access. Knowledge module 558 provides a number of data requests to data module 554, via interface 562.

[0177] Data module 554 submits the data requests provided by the knowledge module 558 to a survey analysis program, and preferably to the RAP survey analysis program. Data module 554 may submit the data requests as a batch RAP command 556, which may execute the RAP requests in a batch mode. The RAP survey analysis program may then provide a number of intermediate RAP reports 561 to data module 554. The intermediate RAP reports 561 may be in an ASCII text format. Data module 554 may then parse the intermediate RAP reports 561 into a number of data objects, and may send the resulting data objects to knowledge module 558 via interface 570.

[0178] Knowledge module 558 uses the parsed intermediate reports and/or data objects to calculate and/or assemble the desired reports. Preferably, this is accomplished by executing a number of rules in a rules-based expert system. The desired reports are then provided to interface module 552 via interface 572, and stored as a number of analysis reports 576.

[0179] It is contemplated that knowledge module 558 may determine a number of result caveats, preferably by executing a number of rules on a rules-based expert system. The result caveats may indicate if the desired results should be viewed with caution. For example, the result caveats may indicate if the results provided to the user are based on a statistically insignificant sample size.

[0180] FIG. 21 is a flow diagram showing the general execution of a request in accordance with the preferred embodiment of the present invention. Initializing block 600 clears the transaction viewer and attaches a first user request. The first user request is sent to the knowledge module for verification. Verifying block 602 verifies the user request and appends any appropriate request caveats to a list of request caveats. If the user request is determined to be appropriate, control is passed directly to executing block 606. If the user request is not determined to be appropriate, appropriate, verifying block 602 asks whether the user request should be executed anyway. If the user indicates that the user request should be executed anyway, control is passed to executing block 606. If, the user indicates that the user request should not be executed anyway, control is passed to viewing block 604.

[0181] Executing block 606 executes the user request, and generates a number of results and result caveats. Control is then passed to viewing element 604. Viewing element 604 displays the user request, the request caveats, the results, and the result caveats.

[0182] FIG. 22 is a low level object model of the preferred interface module. For illustration purposes, each transaction 622 may include a query class 624 and a result class 626. It should be understood, however, that in the preferred embodiment, the query class 624 and the result class 626 are not actually provided, and are only shown for illustrative purposes. The query class 624 may include a SWOT entity class 628, a request class 630 and a number of request caveat objects 632. Preferably, the request caveat objects 632 are built by the knowledge module, and can be viewed by the user.

[0183] The request class 630 can be saved, independent of any system settings or parameters to support the scripting feature of the present invention. The request class 630 may include a simple RAP request class 634 and/or a complex request class 636. The simple RAP request class 634 identifies specific RAP reports that are requested. The complex request class 636 identifies a specific SWOT request or survey analysis request. The result class 622 includes result data class 638 and a number of result caveats objects 640. The results are preferably built using data objects provided to the knowledge module from the data module.

[0184] FIG. 23 is a composite object model for the preferred embodiment of the knowledge module. The knowledge module is shown at 650, and may include query analysis class 652, KM_AGENDA objects 654, inference network class 656, KM_REQUESTS class 658 and KM_RESULTS class 660. Query analysis class 652 may include query execute class 662 and a query validation class 664. Query execute class 662 builds the result objects, and sends the result caveats to the interface module. Query validation class 664 builds the request caveats and sends them to the interface module. KM_AGENDA objects 654 may store agenda objects which control the rule driven report generation and rule-set application by the inference network 656.

[0185] The inference network class 656 may include rule-set objects 668 and explain objects 670. Preferably, the inference network class contains a number of multi-valued slots (lists), where each slot corresponds to a rule set. The list of rules to be executed in each instance is stored in the slot name, and the knowledge module determines which rule sets to invoke. A list of rule sets for each type of SWOT analysis is shown in FIG. 51. A list of rules for each rule set is shown in FIGS. 52A-52C. A description of each rule is shown in FIGS. 55A-550. Finally, inference network class 656 preferably stores a number of explanation objects 670 which correspond to each request. The explanation objects 670 are returned to the interface module to support the “explain” function.

[0186] KM_REQUEST class 658 preferably stores a number of objects that correspond to the knowledge module requests. The request objects are sent to the data module for processing. KM_RESULTS class 660 includes a SWOT results class 672, a number of RAP result objects 674 and a survey results class 676. SWOT results class 672 stores strength objects 686, weakness objects 688, opportunity objects 690 and threat objects 692. The executed queries indicated at 662 provide the strength objects 686, weakness objects 688, opportunity objects 690 and threat objects 692, as indicated by line 694.

[0187] The survey results class 676 includes correlation objects 678, GAP objects 680, MGAP objects 682 and distribution objects 684. The correlation objects 678, GAP objects 680, MGAP objects 682 and distribution objects 684 are created as a result of RAP requests made to the survey database. The strength objects 686, weakness objects 688, opportunity objects 690, threat objects 692, correlation objects 678, GAP objects 680, MGAP objects 682 and distribution objects 684 can be returned to the interface module.

[0188] FIG. 24 is a diagram showing a functional model of a SWOT (Strength/Weakness/Opportunity/Threat) analysis 700. SWOT analysis 700 is initiated by a complex query 702. The complex query 702 indicates whether a strength, weakness, opportunity or threat analysis is to be performed. SWOT analysis 700 determines the type of analysis to be performed, analyzes the request, and provides a number of request caveats 724. This is preferably done by executing a number of rules in a rules-based expert system.

[0189] SWOT analysis 700 then provides a number of report requests to the data module, as shown at 720. The data module 720 may generate a distribution report 704, a MGAP report 706 and/or a correlation report 708. The data module may either execute the requests directly on a survey database, or provide the request to a survey analysis program, as described above. SWOT analysis 700 then reads the distribution report 704, MGAP report 706 and/or correlation report 708 returned as data objects from data module. From these reports, SWOT analysis 700 generates the appropriate results as shown at 710 through 718, depending on the type of analysis performed. For example, if the request provided by the complex query 702 is a strength analysis, SWOT analysis 700 provides strength results 712. In addition to providing the results, SWOT analysis 700 may provide a number of result caveats, as shown at 724. Further, SWOT analysis 700 may indicate which rules and rule sets were executed during the SWOT analysis as shown at 722.

[0190] FIG. 25 is a diagram showing a functional model of a strength analysis. The complex query is formed at 730 and provided to analyze request block 732. Analyze request block 732 determines the type of request that has been provided. In the illustrative diagram, analyze request block 732 determines that the request provided by complex query 730 is a strength request, as indicated at 736. Thereafter, analyze request block 732 creates the appropriate result objects, as shown at 734.

[0191] Generate strength block 736 requests the appropriate reports from data module 738. Data module 738 either directly access the survey database, or provides a request to a survey analysis program. In either case, the MGAP report 740, distribution report 742 and correlation report 744 may be created. The generate strength block 736 reads the MGAP report 740, distribution report 742 and/or correlation report 744, and determines the requested strengths. The generate strength block 736 then stores the strengths under the proper strength instance, as shown at 746. The strength data is then provided to send results block 748, which sends the results to results block 750.

[0192] Generate strength block 736 also checks the survey data at 756 and provides a number of result caveats, as shown at 758. Finally, generate strength block 736 creates explanation objects, as shown at 752, and provides the rules and rule sets that were executed during the strength analysis including comments therefor, as shown at 754.

[0193] FIG. 26 is a diagram showing a functional model of a weakness analysis. The complex query is formed at 770 and provided to analyze request block 772. Analyze request block 772 determines the type of request that has been provided. In the illustrative diagram, analyze request block 772 determines that the request provided by complex query 770 is a weakness request, as indicated at 776. Thereafter, analyze request block 772 creates the appropriate result objects, as shown at 774.

[0194] Generate weakness block 776 requests the appropriate reports from data module 778. Data module 778 either directly access the survey database, or provides a request to a survey analysis program. In either case, the MGAP report 780, distribution report 782 and correlation report 784 may be created. The generate weakness block 776 reads the MGAP report 780, distribution report 782 and/or correlation report 784, and determines the requested. weaknesses. The generate weakness block 776 then stores the weaknesses under the proper weakness instance, as shown at 786. The weakness data is then provided to send results block 788, which sends the results to results block 790.

[0195] Generate weakness block 776 also checks the survey data at 796 and provides a number of result caveats, as shown at 798. Finally, generate weakness block 776 creates explanation objects, as shown at 792, and provides the rules and rule sets that were executed during the weakness analysis including comments therefor, as shown at 794.

[0196] FIG. 27 is a diagram showing a functional model of an opportunity analysis. The complex query is formed at 800 and provided to analyze request block 802. Analyze request block 802 determines the type of request that has been provided. In the illustrative diagram, analyze request block 802 determines that the request provided by complex query 800 is an opportunity request, as indicated at 804. Thereafter, analyze request block 802 creates the appropriate result objects.

[0197] Generate opportunity block 804 requests the appropriate reports from the data module. For an opportunity request, this includes both a strength analysis 806 for selection-1 and a weakness analysis 808 for selection-2. Selection-1 and selection-2 may correspond to an selection set including selected vendors, geographic regions, products, years, etc.

[0198] The generate opportunity block, 804 reads the results from the strength analysis 806 for selection-1 and from the weakness analysis 808 for selection-2, and determines the requested opportunities for selection-1. The generate opportunity block 804 then stores the opportunities under the proper opportunity instance, as shown at 810. The opportunity data is then provided to send results block 812, which sends the results to results block 814.

[0199] Generate opportunity block 804 also checks the survey data at 820 and provides a number of result caveats, as shown at 822. Finally, generate opportunity block 804 creates explanation objects, as shown at 816, and provides the rules and rule sets that were executed during the opportunity analysis including comments therefor, as shown at 818.

[0200] FIG. 28 is a diagram showing a functional model of a threat analysis. The complex query is formed at 830 and provided to analyze request block 832. Analyze request block 832 determines the type of request that has been provided. In the illustrative diagram, analyze request block 832 determines that the request provided by complex query 830 is a threat request, as indicated at 834. Thereafter, analyze request block 832 creates the appropriate result objects.

[0201] Generate threat block 834 requests the appropriate strength and weakness analysis, as described above. For a threat request, this includes both a weakness analysis 836 for selection-1 and a strength analysis 838 for selection-2. Selection-1 and selection-2 may correspond to any selection set including selected vendors, geographic regions, products, years, etc.

[0202] The generate threat block 834 reads the results from the weakness analysis 836 for selection-1 and from the strength analysis 838 for selection-2, and determines the requested threats for selection-1. The generate threat block 804 then stores the threats under the proper threat instance, as shown at 840. The threat data is then provided to send results block 842, which sends the results to results block 844.

[0203] Generate threat block 834 also checks the survey data at 850 and provides a number of result caveats, as shown at 852. Finally, generate threat block 834 creates explanation objects, as shown at 846, and provides the rules and rule sets that were executed during the threat analysis including comments therefor, as shown at 848.

[0204] FIG. 29 is a diagram showing a functional model of a survey validation analysis. A survey validation request is a form of a complex request described above, and is shown at 870. Using the survey validation request, block 872 requests the appropriate reports from the data module. The data requests are provided to the data module, as shown at 874. In a preferred embodiment, the survey validation request causes the data module to submit the appropriate data requests to the RAP program, thereby generating distribution reports 876, MGAP reports 892, Correlation reports 894, and/or other reports 900.

[0205] The distribution reports 876 are provided to distribution checks block 878 and to demographic checks block 880. Distribution checks block 878 checks the distribution of the respondents to various survey questions. For example, the distribution checks block 878 may check the distribution of survey respondents among several customer groups including financial customers, government customers, commercial customers, etc. The results are provided to results block 890. If the number of respondents from any group is outside of a predetermined range, the distribution checks block 878 may provide a result caveat to caveats block 891.

[0206] The demographic checks block 880 may check the distribution of respondents on a geographical basis. The results are provided to results block 890. If the number of respondents from any geographic area is outside of a predetermined range, the distribution checks block 878 may provide a result caveat to caveats block 891.

[0207] The MGAP reports 892 are provided to paired response check block 896. Paired response check block 896 checks the gaps between importance and satisfaction for various survey questions. The results are provided to results block 890. If the gaps are outside of a predetermined range, the paired response check block 896 may provide result caveats to caveats block 891.

[0208] The statistics reports 894 are also provided to paired response check block 896. Using the statistics reports 894, paired response check block 896 may determine if only one answer is provided for a paired response. For example, the paired response check block 896 may determine if the respondent only answered the importance question, but not the satisfaction question for a paired response type question in a survey. The results are provided to results block 890. If only one answer is provided for a paired response, the paired response check block 896 may provide a result caveat to caveats block 891.

[0209] Finally, other reports block 900 provides other report to additional surveys check block 902. Like above, additional survey check block 902 provides the results to results block 890, and further provides a number of result. caveats to caveats block 891.

[0210] FIG. 30 is a diagram showing a functional model of the preferred data module. The data module is illustrated at 930 and receives requests from both the knowledge module 932 and the interface module 952.

[0211] As indicated above, interface module 952 may display a listing of generic questions, which may be selected by the user when forming a user request. To determine the appropriate generic questions for a selected survey, the interface module may access the correlation data 946 via the data module 930, and retrieve the generic questions that correspond to the actual questions in the selected survey.

[0212] Likewise, the interface module 952 may allow a user to select one of a number of answers to a selected question. The available answers are obtained by accessing the question/answer (MQA) data block 950 via data module 930, as shown. The question/answer data block 950 includes the available answers to each question in a survey.

[0213] Once a user request is formed, the request is provided to knowledge module 932. Knowledge module 932 may identify the actual question numbers and the appropriate surveys that correspond to the generic questions provided in the user request. This is preferably accomplished by accessing correlation data 946 via data module 930, as described above.

[0214] The knowledge module 932 assembles a number of requests, and provides the requests to the data module 930. Preferably, the requests identifies the appropriate data elements within the RAP report data 940 by specifying the actual question numbers and the corresponding survey.

[0215] It is contemplated that some requests may only request RAP reports from the RAP program. These requests are specified by selecting the RAP command mode in the analysis control window, shown for example in FIG. 8. For these requests, the data module 930 may provide the requests directly to RAP activity manager 936 via interface 931. The RAP activity manager may execute the requests, preferably in a batch mode as shown at 944. The resulting RAP reports may then be provided back to data module 930 via interface 933, and returned to the interface module 952 via knowledge module 932.

[0216] For complex requests, knowledge module 932 may determine which RAP requests are required to derive the desired results. The knowledge module 932 sends the resulting RAP requests to data module 930 When data module 930 receives the RAP requests from knowledge module 932, data module 930 provides the requests to report block 934. Report block 934 provides the RAP requests to RAP activity manager 936.

[0217] RAP activity manager 936 accesses the RAP report data 940 and provides the requested RAP reports to RAP activity manager 936. Preferably, the RAP activity manager 936 provides a batch RAP request 944 to RAP for batch processing. The RAP activity mianager 936 receives the requested reports from RAP, and provides the RAP report data to the report block 934. Report block 934 then preferably parses the RAP reports, which are typically in an ASCII format, into a number of requested data objects. The request data objects are then returned to knowledge module 932 via data module 930 for further processing.

[0218] FIG. 31 is a diagram showing a consolidated low level object model of the preferred data module. RAP activity manager 936 includes RAP batch interface 960 and RAP file manager 962. RAP batch interface 960 allows batch execution of RAP requests, as shown at 944. The output of the batch execution of RAP requests is provided to RAP report data store 961. RAP file manager 962 may read the RAP report data store 961 and provide the results to RAP report parser 964. RAP activity manager 936 may control RAP report parser 964. RAP report parser 964 parses the survey data 966 into a number of data objects. For example, RAP report parser 964 may provide distribution objects 968, correlation objects 970, statistics objects 972, direct objects 974 and GAP objects 976, as shown.

[0219] As indicated above, a correlation table manager 978 may access a correlation table 980 to provide a correlation between questions in various surveys. Correlation table 980 may include question strings 982 and survey translation objects 984.

[0220] FIG. 32 is a diagram showing a functional model of a preferred correlation table manager. Correlation table manager 1040 receives correlation data requests 1042 from the data module. Correlation table manager 1040 provides the survey correlation request to correlation table 1044. Correlation table 1044 is updated from the correlation table data store 1046 whenever a new correlation table is generated, for example, whenever a new survey is received and incorporated into the system.

[0221] A correlation data request 1042 may provide a generic question, and may request an actual question number in a selected survey. As indicated above, this type of request is often provided by the knowledge module. The correlation table manager provides the request to correlation table 1044. Correlation table provides a text-to-question correlation request to survey-xlation block 1052. In response, survey-xlation block 1052 provides the corresponding question number to correlation table 1044. The corresponding question number is then provided to the requestor via correlation table manager 1040.

[0222] Another type of correlation data request 1042 is a generic text request. It is often desirable for both the data module and knowledge module to translate from a question number to a corresponding generic question. For example, the data module may receive a number of RAP reports in response to a RAP request. The RAP reports only include the question text from a specific survey. Before transferring the RAP reports, it may be desirable to translate the question text in the RAP reports to a corresponding generic question string.

[0223] To do this, the data module may provide a correlation data request to correlation table manager 1040. The correlation table manager 1040 provides the request to correlation table 1044. Correlation table 1044 provides a generic question request to question string block 1050. The generic question request may identify both the question number and the specific survey. In response, the question string block 1050 may provide the corresponding generic question string to correlation table 1044. The corresponding generic question string may then be provided to the requestor via correlation table manager 1040.

[0224] Likewise, it may be desirable for the knowledge module to translate from a question number to a generic question string. For example, the knowledge module may perform a strength type analysis. The data module may parse the resulting RAP reports into a number of data objects, and provide the resulting data objects to the knowledge module. It may be desirable for the knowledge module to translate the question text within the reports to the corresponding generic question string before providing the results to the interface module. This translation operates similar to that described above with respect to the data module.

[0225] FIG. 33 is a table showing the execution of an illustrative direct RAP request, with a problem detected in the request. The user first enters a direct RAP command via the interface module. The interface module creates a simple RAP object for further analysis. The simple RAP object is sent to the knowledge module for verification. The knowledge module receives the simple RAP object and uses rules to check for problems in the request. Problems that are detected include syntax errors or requests that are determined to be not appropriate. The problems are noted and a number of request caveats provided. The request caveats are sent back to the interface module. The interface module appends the request caveats to a listing. The user then views the request caveats.

[0226] FIGS. 34A-34B show the execution of an illustrative direct RAP request, when no problems are detected in the request or when the user indicates that the request should be executed despite the detected problem. The user enters the RAP command, and the interface module accepts the RAP command. The interface module creates a simple RAP object for further analysis. The simple RAP object is then sent to the knowledge module for verification.

[0227] The knowledge module receives the simple RAP object and uses rules to check for problems in the request. The knowledge module notes any problems, and updates the request caveats. The request caveats are sent back to the interface module. The interface module then appends the request caveats to a listing.

[0228] The knowledge module also sends the RAP command to the data module. The data module receives the RAP command object, and creates batch commands therefrom. The data module executes the batch command and returns the direct RAP objects. The knowledge module previews the RAP results and sends the RAP results to the interface module. The interface module appends the results, and displays the results to the user. The user then views the results, and the result and request caveats.

[0229] FIGS. 35A-35B show the execution of an illustrative strength/ weakness/opportunity/threat analysis, when problems are detected in the request. The user selects one or more surveys and request a strength, weakness opportunity or threat analysis. The interface module accepts and checks the request. The interface module also creates a complex query object and sends the complex query object to the knowledge module.

[0230] The knowledge module receives the complex query object, and evaluates the analysis inputs. The knowledge module determines the type of reports needed to validate the request, and creates the knowledge module request for the data module.

[0231] The data module receives the knowledge module request object and checks for existing reports/objects that correspond to the knowledge module request. If not found, the data module runs the appropriate RAP report, and generates the corresponding objects. In either case, the data module returns the requested objects to the knowledge module.

[0232] The knowledge module examines the returned objects and generates request caveats based on knowledge module rules. The knowledge module also creates a number of result caveat objects and sends them to the interface module. The interface module receives the results, request caveats, and result caveats, and displays them to the user.

[0233] FIGS. 36A-36C show the execution of an illustrative strength or weakness analysis, when no problems are detected in the request or when the user indicates that the request should be executed despite the detected problem. The user selects one or more data survey databases and requests a strength or weakness analysis. The interface module accepts and checks the command, and creates a complex query object. The interface module then sends the complex query object to the knowledge module.

[0234] The knowledge module receives the complex query object, and determines the type of reports needed. For example, the complex query object may indicate to which products or vendors the request is directed. The knowledge module creates knowledge module request objects for the data module and sends the knowledge module request objects to the data module.

[0235] The data module receives the knowledge module request objects. The data module checks existing reports and objects to determine if these objects have been previously generated. If not, the data module runs the appropriate RAP reports and generates appropriate objects. In either case, the data module returns the appropriate results objects to the knowledge module.

[0236] In a preferred embodiment, the result objects includes MGAP objects. The knowledge module looks through the MGAP objects and determines which of the MGAP objects are strengths or weaknesses based on a number of rules. The knowledge module then creates a result object stream and sends it to the interface module. The interface module receives the appended result and displays it to the user. The user receives the strength or weakness analysis results, and the corresponding request and result caveats, if any.

[0237] FIGS. 37A-37C show the execution of an illustrative threat or opportunity analysis, when no problems are detected in the request or when the user indicates that the request should be executed despite the detected problem. The user selects the appropriate survey information. This may include one or more surveys, vendors, products, etc. The user also selects an opportunity or threat analysis type. The interface module accepts and checks the command and creates a complex query object. The interface module sends the complex query object to the knowledge module.

[0238] The knowledge module receives the complex query object and determines the type of reports needed based on the products, vendors, etc. that are selected. The knowledge module creates multiple knowledge module requests for the data module based on vendor and product knowledge (e.g. rules). The knowledge module then sends the multiple knowledge module requests to the data module.

[0239] The data module receives the knowledge module request objects and checks existing reports and objects to see if these objects have been generated by a previous request. If not found, the data module runs the appropriate RAP reports and generates the appropriate objects. In either case, the data module returns the result objects to the knowledge module.

[0240] In a preferred embodiment, the result objects include MGAP objects. The knowledge module looks through the MGAP objects and determines which are strengths or weaknesses based on predefined rules. The knowledge module compares these strengths and weaknesses, based on the selected vendors and/or products, and determines which are threats and which are opportunities. The knowledge module creates a result object stream and sends the result object stream to the interface module.

[0241] The interface module receives the appended results and displays the results to the user. The user receives the threats or opportunities analysis results, and the corresponding request and result caveats, if any.

[0242] FIG. 38 is a table showing the execution of an illustrative survey validation request, with request problems detected. The user enters a survey validation request. The interface module accepts the survey validation request, and creates survey validation objects for further analysis. The survey validation objects are sent to the knowledge module for verification.

[0243] The knowledge module receives the survey validation objects, and uses a number of rules to check for problems in the request. The knowledge module notes any problems and updates the request caveats. The knowledge module sends the request caveats back to the interface module. The interface module appends the request caveats, and the user views the appended survey validation request caveats.

[0244] FIGS. 39A-39B show the execution of an illustrative survey validation request, when no query problems are detected or when the user indicates that the request should be executed despite the detected problems. The user enters the appropriate survey information and requests a survey validation. The survey validation request is a form of a complex request. The appropriate survey information may include the selection of a particular survey. The interface module accepts and checks the survey validation request, and creates a survey validation object. The survey validation object is sent to the knowledge module.

[0245] The knowledge module receives the survey validation object and determines the type of reports needed based on the products, vendors, etc. The knowledge module then creates a number of data request objects, and provides the data request objects to the data module.

[0246] The data module receives the data request objects and checks existing reports/objects to determine if the requested reports/objects were already generated by a previous request. If not found, the data module runs the appropriate RAP reports and generates the appropriate result objects. In either case, the data module returns the appropriate result objects to the knowledge module.

[0247] The knowledge module looks through the result objects that were returned, and used rules to generates survey validation results. The knowledge module creates a result objects stream and sends the result object stream to the interface module. The interface module receives the appended results, and displays the results to the user.

[0248] FIG. 40 is a table showing the execution of an illustrative SWOT parameter update using the SWOT editor of FIG. 17. The user enters SWOT and/or analysis parameters via the SWOT editor. The interface module accepts and checks the parameters entered by the user. The interface module informs the user of nonacceptable parameter values, if any. The interface module then allows the user to store or retrieve these parameters upon request. The user receives an updated display.

[0249] FIG. 41 is a flow diagram showing an illustrative method of the present invention. The algorithm is entered at element 1060 and control is passed to element 1062. Element 1062 accepts a user request. Control is then passed to element 1064. Element 1064 processes the request by executing selected ones of a number of predefined rules in a rules based expert system to provide a survey data request. Control is then passed to element 1066. Element 1066 accesses the survey database in accordance with the survey data request, and provides a number of selected survey data elements. Control is then passed to element 1068. Element 1068 processes selected ones of the selected survey data elements by executing selected ones of a number of predefined rules in a rules based expert system. Control is then passed to element 1070. Element 1070 provides a result. Control is then passed to element 1072, wherein the algorithm is exited.

[0250] FIG. 42 is a flow diagram showing another illustrative method of the present invention. The algorithm is entered at element 1200 and control is passed to element 1202. Element 1202 provides a user request. Element 1204 checks the user request to determine if it is a valid request. If the request is not valid or otherwise questionable, a request caveat may be provided. Control is passed to element 1206. Element 1206 determines which of the data elements in the survey data base are required to derive a response to the user request by, at least in part, executing a number of predefined rules in a rules-based expert system, thereby resulting in a number of selected data elements. Control is then passed to element 1208.

[0251] Element 1208 reads the selected data elements from the survey data base, and passes control to element 1210. Element 1210 determines the response to the request at least in part by executing a number of predefined rules in a rules-based expert system. Control is then passed to element 1212. Element 1212 determines if the survey database has a sufficient number of selected data elements to derive a statistically significant response to the user request. Control is then passed to element 1214, wherein the algorithm is exited.

[0252] FIG. 43 is a flow diagram showing the method of FIG. 42, with elements 1206, 1208 and 1210 of FIG. 42 replaced with elements 1220 and 1222 of FIG. 43. Referring back to FIG. 42, after element 1204 checks the user request, element 1220 of FIG. 43 assembles a number of data requests using the rules-based expert system. Element 1220 further submits the number of data requests to a data module, and provides a number of requested intermediate reports to the rules-based expert system. Element 1222 uses the number of requested intermediate reports to calculate a number of entries in one or more reports that are provided to the user. Control is then passed to element 1212 of FIG. 42. It is contemplated that element 1220 and element 1222 may be combined, wherein the data module may access the survey database and provides one or more reports that are then provided to the user.

[0253] FIG. 44 is a flow diagram showing another variation of the method shown in FIG. 42, with element 1206 of FIG. 42 is replaced with element 1228 of FIG. 44. Element 1228 determines an analysis type of the user request, and determines which of the data elements in the survey database are required to derive a response to the analysis type of the user request. This is done, at least in part, by executing a number of predefined rules in a rules-based expert system, thereby resulting in a number of selected data elements.

[0254] FIG. 45 is a flow diagram showing an illustrative method for providing ordered processing in a rules based expert system. The algorithm is entered at element 1230 and control is passed to element 1232. Element 1232 submits a first rule to an inference engine, wherein the first rule generates a number of first agenda objects within an object oriented database. Control is then passed to element 1234. Element 1234 executes the number of first agenda objects in a predetermined order. Control is then passed to element 1236, wherein the algorithm is exited.

[0255] FIG. 46 is a flow diagram showing another illustrative method for providing ordered processing in a rules based expert system. The algorithm is entered at 1240, and control is passed to element 1242. Element 1242 submits a number of rule sets to an inference engine, wherein each of the rule sets generate a number of agenda objects within an object oriented database. Preferably, each of the agenda objects specify a data request and a number of rules. Control is then passed to element 1244.

[0256] Element 1244 processes a first one of the number of agenda objects, including executing the corresponding data request, thereby providing a number of first data elements. Thereafter, the rules associated with the first agenda object are executed using the inference engine. Preferably, the number of rules specify further processing of the first data elements. Control is then passed to element 1246.

[0257] Element 1246 processes a second one of the number of agenda objects using the inference engine after the first one of the number of agenda objects is processed by element 1244. The second agenda object is processed in a similar manner as the first agenda object described above. Control is then passed to element 1248, wherein the algorithm is exited.

[0258] This method is useful for providing ordered processing in a rules based expert system. One such example of ordered processing is in the survey validation procedure of the present invention. The Knowledge Module preferably includes an object called KM_agenda (see FIG. 54D and FIG. 55D), that serves as a sub-script mechanism for multi-stage tasks. The KM_agenda instances include all of the slots of an IM request object, a rule set list slot, and several flags to indicate whether to report the data directly to the results window and/or analyze the data and return those results. Rule set execution can be iterative. That is, an agenda object may create additional agenda objects, which must then be processed.

[0259] Survey Validation uses the agenda concept to perform a sequence of checks and operations on a selected survey. In a preferred embodiment, Survey Validation processing includes the steps of:

[0260] 1. Execute rule set RS_init_validate_gen, which includes rules v-gen-1 . . . v-gen-n. Each of these rules creates an agenda object, populates the request fields, and sets the appropriate rule set list names. This includes specifying the type of analysis, the RAP reports to generate, whether or not to directly report the RAP results or just the rule-based conclusion upon completion of the task; and

[0261] 2. Execute each agenda object in a predetermined order, and apply the list of rule sets associated with the agenda instance.

[0262] Using this mechanism, adding additional checks to Survey Validation can be accomplished by adding an appropriate rule to RS_init_validate_gen, wherein the rule includes code to create and populate a KM_agenda object with survey data (object variables) and appropriate rule sets. The rule sets can be new or existing, and may operate on the data generated by the analysis specified in the agenda object. The Validation command may then process the additional agenda objects automatically via the inference engine.

[0263] FIG. 47 is a flow diagram showing an illustrative method for providing scripting of requests. The algorithm is entered at element 1270, wherein control is passed to element 1272. Element 1272 provides a system for processing a number of user requests, wherein the user requests are submitted to an inference engine of a rules-based expert system. The rules-based expert system has a number of system settings that effect the processing of the rules. Element 1274 saves selected ones of the user requests independent of the system settings of the rules-based expert system, thereby forming a user request script. Control is then passed to element 1276. Element 1276 sets the system settings. Control is then passed to element 1278. Element 1278 reads the user request script and processes the corresponding user requests therein using the rules-based expert system in accordance with the new system settings. Control is then passed to element 1280, wherein the algorithm is exited.

[0264] FIG. 48 is a flow diagram showing an illustrative method for increasing the performance of a survey database analysis system by using previously generated results. The algorithm is entered at element 1290 and control is passed to element 1292. Element 1292 sequentially executes a number of user requests on the database, wherein each of the user requests causes a number of data requests to be provided to the database, and each of the number of data requests provides a result. Element 1294 determines the data requests that are required for a selected user request. Element 1296 determines if a first data request that is required for the selected user request has been previously executed. Element. 1298 uses selected results of a previously executed data request during the execution of the first data request if it is determined that the first data request has been previously executed. Control is then passed to element 1300, wherein the algorithm is exited.

[0265] FIG. 49 is a flow diagram showing an illustrative method for identifying rules that were executed in a rules based expert system. The algorithm is entered at 1310, wherein control is passed to element 1312. Element 1312 executes a first execution command run in a rules-based expert system. Control is then passes to element 1314. Element 1314 stores the rule names of the rules that are applied during the first execution command run. Control is then passed to element 1316. Element 1316 executes a second execution command run in the rules-based expert system. Control is then passed to element 1318. Element 1318 stores the rule names of the rules that are applied during the second execution run. Control is then passed to element 1320. Element 1320 provides the stored rule names for both the first and second execution command runs to the user upon request. Control is then passed to element 1322, wherein the algorithm is exited.

[0266] FIG. 50 is a flow diagram showing an illustrative method for detecting when a rules based expert system is stuck in a loop. The algorithm is entered at element 1330, wherein control is passed to element 1332. Element 1332 maintains a count for each of the number of rules. The count indicating the number of times the corresponding rule is executed by the rules-based expert system. Control is then passed to element 1334. Element 1334 interrupts the rules-based expert system if the count corresponding to selected rules exceeds a predetermined value. Control is then passed to element 1336, wherein the algorithm is exited.

[0267] FIG. 51 is a table showing illustrative analysis types, and the corresponding rule sets that are associated therewith. In the left column, the preferred complex analysis types are shown. In the right column, the preferred rule sets are identified for each analysis type. When a particular analysis type is requested, the knowledge module executes each of the corresponding rule sets in the order indicated.

[0268] FIGS. 52A-52C show a table of preferred rule sets, and the corresponding rules that are associated therewith. In the left column, a number of rule set-names are shown. In the right column, the preferred rules are identified for each rule set. When a particular rule set is executed by the knowledge module, the corresponding rules are executed by the inference engine as required.

[0269] FIG. 53A-530 show a table of the preferred rules, along with corresponding rule comments. In the left column, a number of rule names are shown. In the right column, the corresponding rule comments are shown. The rule comments describe the action performed by the corresponding rule.

[0270] FIG. 54A-54F shows an illustrative listing of the object oriented database before any analysis runs are executed. FIG. 55A-55G shows an illustrative listing of the object oriented database before after a strength analysis type is executed. When comparing FIG. 54D and FIG. 55D, it can be seen that a number of strength results are generated in the KM/KM_results/SWOT_results/strength class.

[0271] Having thus described the preferred embodiments of the present invention, those of skill in the art will readily appreciate that the teachings found herein may be applied to yet other embodiments within the scope of the claims hereto attached.