JP5105056B2 - Questionnaire analysis system, questionnaire analysis method and program - Google Patents

Description

  The present invention relates to a questionnaire system and method for separating, analyzing, and analyzing differences in intention, sensitivity, and opinion caused by a questionnaire method, a questionnaire target person, and a questionnaire target.

  In recent years, with the diversification and fragmentation of human sensibilities, individual sensibilities and expressions tend to widen from individual to individual.

  As a related technique, a system that extracts and uses personal values has been proposed.

  For example, in Patent Document 1, the results of a questionnaire enumerating vocabulary are statistically processed, the internal values and the target evaluation values are quantified by factors, and the internal values and the target evaluation values of the subject are determined. It is obtained as a linear model, and the internal values and target evaluation values of the subject are obtained using similarity and estimation. Furthermore, information is provided to the subject using the obtained internal values and target evaluation values.

  In addition, as related research, in Non-Patent Document 1, orthogonal procratic analysis is performed on questionnaire results obtained by the SD (Semantic Differential) method, and attention is paid to the similarity of the data structure of each subject obtained. The contents for classifying the subject are described.

JP 2001-344375 A Noboru Kuroda et al. Study on analysis of individual impression evaluation by clustering method considering distribution structure, Japan Society for Kansei Engineering 2006 Spring

  In related technologies and research, it is common to conduct a questionnaire and perform analysis and analysis using the collected questionnaire results. In addition, it is known that it is difficult to apply a conventional questionnaire to extract differences based on ideas such as disagreements, knowledge levels, and emotions at that time. That is, there is a problem that ideas such as differences in opinion and sensitivity, knowledge level, and emotions at that time cannot be extracted well. In other words, there is a problem that it is not possible to separate differences in opinions and sensitivities to the question object itself and differences in expression and interpretation during image transmission.

  As a specific example, when communicating in a situation where both parties do not draw a concrete image of the content or do not draw a full image, the information transmitted by communication includes the difference in opinion and sensitivity to the target , Including differences in image representation and interpretation.

  In other words, the related questionnaire analysis method cannot take into account the breakdown of the evaluation of the evaluation target even if the technology and method of extracting the related individual's sensibility and opinions are used, and both sides can draw a concrete image. Effective processing cannot be performed in some situations.

  The object of the present invention is to solve the above-mentioned problem, to ask questions by a similar questionnaire for a plurality of objects in addition to the target evaluation object, and to have a data structure in the score data regarding each object of each subject, Information processing equipment and behavior analysis programs that enable the extraction of evaluation patterns between subjects from questionnaire data, and the analysis and extraction of differences in the sense of interpretation and interpretation of ambiguous information on products and services that are not embodied Is to provide.

  Another object of the present invention is to perform a similar questionnaire question on a plurality of subjects that present information that is sensitive to the evaluation as a marker in addition to the target assessment subject, By providing the score data for the subject with a data structure, obtain the data structure of the grade data with a small difference in opinion and sensitivity, and compare it with the data structure of the grade data for the target subject. And information processing devices that can analyze and extract evaluations on subjects' classification and ambiguous services that are specialized only in the sensibility aspect. The purpose is to provide a program for behavior analysis.

The questionnaire analysis system according to the present invention is a questionnaire analysis system for conducting a questionnaire on a plurality of subjects for a subject to be evaluated, and collecting and analyzing the results of the questionnaire. A control unit that performs information processing, a questionnaire result information input unit that receives answers to a questionnaire regarding a target to be evaluated presented to a plurality of subjects and a target with a small difference in interpretation between the subjects, and the questionnaire result information input unit A storage unit that stores and accumulates various types of information including information on answers to the questionnaire acquired in step (a), and the control unit includes a target to be evaluated and a target to be a plurality of the markers when performing the questionnaire To a plurality of subjects, and a pair to be the target to be evaluated and the plurality of markers A questionnaire result of multiple persons obtained by asking a question with the same content as the elephant is obtained via the questionnaire result information input unit, and score data is extracted from the score questionnaire result and recorded in the storage unit Score data relating to a plurality of markers recorded from the storage unit as a marker target group for each subject as a marker target group, and interpretation of each marker from the characteristics of the score data of the marker target group Interpretation difference extraction means for extracting features that show a relative relationship with other subjects, and score data related to the target group including the target to be evaluated recorded from the storage unit for each subject. In addition to extracting and processing the features that showed a relative relationship with other subjects regarding how to assign scores for each subject, The difference between the relative characteristics of the subject regarding how to assign a score to other subjects and the relative characteristics of the interpretation of the subject extracted by the interpretation difference extraction means with respect to other subjects is compared. And a sensitivity difference extraction means for separating a difference in sensitivity evaluation relative to other subjects to the subject to be evaluated, which is reflected in a score questionnaire result including sensitivity and interpretation. The interpretation and sensitivity of a given subject are classified so as to be relatively comparable between subjects.

  ADVANTAGE OF THE INVENTION According to this invention, the questionnaire analysis system, method, and program which can analyze the interpretation and sensitivity with respect to evaluation object by acquiring the result of the questionnaire of evaluation object, and analyzing and analyzing can be provided.

  A first embodiment of the present invention will be described with reference to FIGS.

  FIG. 1 is a configuration diagram illustrating a questionnaire system 1 according to an embodiment. The questionnaire system 1 according to the present embodiment is a general computer system, and has a configuration as shown in FIG.

  That is, the questionnaire system 1 is constructed by a questionnaire server 2 that performs a questionnaire, a questionnaire analysis device 3, and a database 4. Then, the questionnaire server 2 collects the questionnaire by conducting a questionnaire through the network 800 to the questionnaire input device 900 possessed by the subject of the questionnaire to be detected and obtaining an answer. Details will be described later.

  FIG. 2 is a block diagram illustrating functions of the questionnaire system 1 according to the embodiment.

  The questionnaire system 1 includes at least a score data collection unit 10, a score data database 20, an interpretation difference extraction unit 30, a sensitivity difference extraction unit 40, and an analysis result output unit 50. Each unit is distributed and provided in one device or a plurality of devices constituting the questionnaire system 1. In the configuration example shown in FIG. 1, the score data collection unit 10 is a part of the function of the questionnaire server 2, and the score data database 20 is a part of the function of the database 4, and the interpretation difference extraction unit 30, the sensitivity The difference extraction unit 40 and the analysis result output unit 50 are part of the function of the questionnaire analysis device 3.

  The score data collection unit 10 is not limited to collecting data using a network such as the Internet. For example, a medium such as paper or an external memory may be used. At this time, the score data collection unit 10 includes a reading device suitable for the medium.

  The score data collection unit 10 includes a target information presentation unit 101 and a score data input unit 102 as constituent elements, and includes at least one vocabulary to be evaluated as a target for which an evaluation result is to be obtained and a plurality of targets to be markers. The target information of the marker target is presented to each subject, a questionnaire with the same question content is performed on each target, and input of the obtained score data is accepted. It is desirable to use a scale method for the questionnaire.

  The target information presenting unit 101 provides, for each of the evaluation target and the plurality of marker targets, the name of the product or service, its description, a photograph, video, or voice as the target information, and the composite information of each of the subjects. To present.

  The evaluation target does not define the contents of the information to be presented in order to allow ambiguity, but the marker target is selected so that the evaluation deviation by the subject is small. Furthermore, it is desirable to include explanatory text such as effects in the target information so that the evaluation is fixed. Moreover, if the same questionnaire question sentence is applicable, the evaluation target and the marker target do not need to belong to the same category.

  The score data input unit 102 conducts a questionnaire with the same question content on each subject (evaluation target, marker target) for which the target information is presented by the target information presentation unit 101, and scores the results of each subject or his / her agent Data input is accepted and transmitted to the score data database 20. Here, the score data is known to give a score for each questionnaire result by a questionnaire method such as a scale method, and the score information, the total score information, and related items of each questionnaire item. Points information etc.

  The score data database 20 aggregates the score data transmitted from the score data input unit 102 and stores it as data for each subject or each subject. The score data database 20 is not particularly limited as long as it can store a large amount of information. That is, an area of the auxiliary storage device of the information processing apparatus may be used, or an auxiliary storage device installed in a physically separated place via a network may be used.

  The interpretation difference extraction unit 30 includes a marker target group score data extraction unit 301 and a marker target group data structure analysis unit 302 as constituent elements, and features of interpretation between subjects from the difference in the data structure of the score data for the marker target group. To extract.

  The marker target group score data extraction unit 301 extracts and acquires score data related to the marker target group of a plurality of subjects from the score data database 20.

  The marker object group data structure analysis unit 302 analyzes the structure of the score data of each subject acquired by the marker object group score data extraction unit 301. In addition, the characteristics of each subject's information interpretation are analyzed and digitized.

  Most of the differences in the score data structure represent the difference in the subject's interpretation of the subject information presented to the subject because the marker target group makes the subject's sensibility evaluation the same by providing sufficient information .

  There are a plurality of methods for digitizing the information interpretation characteristics of each subject. For example, there is clustering that focuses on the similarity of data structures.

  In addition, there is a difference or deviation measurement between the average (center) data of the subject group to be analyzed and the data of each subject. In addition, you may use methods other than the illustration for digitization of information interpretation.

  The sensitivity difference extraction unit 40 includes a target target group score data extraction unit 401, a target target group data structure analysis unit 402, and a target target evaluation analysis unit 403 as constituent elements. The sensitivity difference extraction unit 40 extracts a difference in sensitivity from the difference between the score data structure for the evaluation target group that is a plurality of target groups including the evaluation target and the score data structure for the marker target group.

  The target object group score data extraction unit 401 is a score on an evaluation target group, which is a plurality of target groups including the evaluation target, of a plurality of subjects identical to the subject whose data is extracted from the score data database 20 by the interpretation difference extraction unit 30. Get the data. The evaluation target group may be a plurality of target groups including at least one or more evaluation targets. When there are a plurality of evaluation targets, the evaluation target group may include only the evaluation targets or may include a marker target.

  The target object group data structure analysis unit 402 analyzes the data structure of the scores of each subject acquired by the target target group score data extraction unit 401, thereby quantifying the rating characteristics of each subject. Since the evaluation target group includes not only the evaluation target but also the marker target, the difference in the data structure of the scores is mixed with the difference in the sensitivity evaluation and the difference in the interpretation.

  There are multiple methods for digitizing the characteristics of scoring for each subject. For example, there is clustering that focuses on the similarity of data structures.

  In addition, there is a difference or deviation measurement between the average (center) data of the subject group to be analyzed and the data of each subject. It should be noted that a method other than the example may be used for digitizing the rating.

  The target object evaluation analysis unit 403 is sensitive to the evaluation object from the difference between the information interpretation characteristics of each subject digitized by the marker object group data structure analysis unit 302 and the scored characteristics of each subject. The differences between different evaluations.

  The analysis result output unit 50 outputs information such as the analysis result of the evaluation to the evaluation object by each subject obtained by the target object evaluation analysis unit 403 and the distribution of each subject.

  In such a configuration, the questionnaire system 1 according to the embodiment makes it possible to analyze the interpretation and sensitivity of the vocabulary of the subject from the questionnaire of the vocabulary to be evaluated.

  Next, the overall operation of the first embodiment of the present invention will be described.

  FIG. 3 is a flowchart illustrating the operation of the questionnaire system 1 according to the embodiment.

  The questionnaire system 1 uses the score data collection unit 10 to present a questionnaire for each of a plurality of marker objects as at least one evaluation object and other objects to be markers. The contents presented in the questionnaire include, for example, the names of products and services, their explanations, photographs, videos, sounds, and composite information. (Step S301).

  The questionnaire system 1 uses the score data collection unit 10 to conduct a questionnaire with the same question content for each subject presented, receives the input of the score data as a result, and transmits it to the score data database 20 ( Step S302). A questionnaire based on a scaling method is desirable.

  The questionnaire system 1 uses the score data database 20 to count the score data related to each target transmitted from the score data input unit 102 and accumulates the data as data for each subject and each target (step S303).

  The questionnaire system 1 uses the interpretation difference extraction unit 30 to extract and acquire score data related to a marker target group of a plurality of subjects from the score data database 20 (step S304).

  The questionnaire system 1 uses the interpretation difference extraction unit 30 to analyze the data structure of the acquired scores of each subject, and digitizes the information interpretation characteristics of each subject (step S305).

  The questionnaire system 1 is a group of a plurality of subjects including a subject's evaluation target of a plurality of subjects who are the same as the subject whose data is extracted by the interpretation difference extraction unit 30 from the score data database 20 using the sensitivity difference extraction unit 40. Score data relating to a certain evaluation target group is extracted and acquired (step S306).

  The questionnaire system 1 uses the sensitivity difference extraction unit 40 to analyze the data structure of the acquired scores of each subject and digitizes the rating characteristics of each subject (step S307).

  The questionnaire system 1 uses the sensibility difference extraction unit 40 to change from the difference between the information interpretation characteristics of each subject quantified by the interpretation difference extraction section 30 and the grading characteristics of each of the quantified subjects to the evaluation target. The difference in the sensibility evaluation is separated (step S308).

  The questionnaire system 1 uses the analysis result output unit 50 to output information such as the analysis result of the evaluation on the evaluation target by each subject obtained by the sensitivity difference extraction unit 40 and the distribution of each subject (step S309).

  Through the series of operations described above, the questionnaire system 1 according to the embodiment can analyze the interpretation and sensitivity of the subject's vocabulary from the vocabulary questionnaire to be evaluated.

  Next, an embodiment that is specifically adapted will be described as an embodiment.

  An example of adaptation will be described using the questionnaire system illustrated in FIG. The questionnaire system 1 illustrated in FIG. 1 is a general computer system.

  The questionnaire system 1 is constructed by a questionnaire server 2 that conducts a questionnaire, a questionnaire analyzer 3, and a database 4. And the questionnaire server 2 collects a questionnaire by implementing a questionnaire in the questionnaire input terminal 900 which the test subject of the questionnaire which is a detection target has, and obtaining an answer.

  The questionnaire input terminal 9 for transmitting a questionnaire result is an information processing terminal such as a general personal computer, a mobile phone, a PHS, a PDA, an IP phone, a FAX, a general subscriber phone, and uses a browser function, a FAX function, and the like. Answer the questionnaire and send information such as the answer result to the questionnaire server via the network 8. Although the network 8 is simply described, it is an Internet line or a public telephone line, and includes a communication infrastructure as necessary.

  An example of this adaptation is to extract a new service that realizes a recycling-oriented society as an evaluation target, and to extract evaluations of the service using a scaling questionnaire. Also, in order to realize a scale method questionnaire, it is connected to a questionnaire input terminal 9 possessed by ordinary people via the network 8, and recruitment of questionnaires, presentation of information, presentation of questionnaire questions, input of answers using the Web Realize the aggregation.

  The score data collection unit 10 and the score data database 20 are included in the questionnaire server 2, and the interpretation difference extraction unit 30, the sensitivity difference extraction unit 40, and the analysis result output unit 50 are included in the questionnaire analysis device 3. 20 is included in the database 4.

  The questionnaire server 2, the questionnaire analysis device 3, and the database 4 include a control unit that performs various types of information processing, a ROM, a RAM, an input / output unit, a storage device, a network interface, and the like, and are connected via a network 8.

  The storage device of each device stores various programs such as an operating system, a questionnaire implementation program that implements various functions, a questionnaire result acquisition program, a questionnaire result analysis program, etc., and is expanded in RAM as necessary. Executed. The storage device can store various information, and information such as score data is recorded in addition to the above-described program.

  The storage device of this embodiment is described using an HDD. It should be noted that similar storage means may be used instead of the HDD. For example, an empty area of flash memory or RAM may be used.

  The hardware configuration of the questionnaire server 2 and the questionnaire analyzer 3 is the same as that of a general information processing apparatus, and will be described using a functional block diagram. The database 4 is a so-called network HDD, and includes an HDD inside, and can communicate with various devices via the network 800. Since the hardware structure and operation are the same as those of a general network HDD, they are omitted. The database 4 is sufficient if the score data database 20 can be realized. Therefore, it is sufficient if the score data can be aggregated and accumulated as data for each subject and each subject.

  FIG. 4 is a block diagram illustrating a schematic configuration of the questionnaire server 2.

  The questionnaire server 2 is a general internet server. As its function, it is possible to collect questionnaire results and the results of the presented questionnaires.

  For questionnaire input terminals 9 such as personal computers and mobile phones capable of displaying HTML, a questionnaire is conducted using HTML. FAX, general subscriber telephones, etc. acquire information using a modem in order to answer a questionnaire without being directly converted into data. Note that the answer may be read using a scanner or the like through another medium such as paper.

  FIG. 5 is a block diagram illustrating a schematic configuration of the questionnaire analysis apparatus 3.

  The questionnaire analysis device 3 is a general computer. As its function, various information can be acquired, processed, and output. It is a so-called personal computer. A program for analyzing the questionnaire is installed.

  The questionnaire system 1 (the questionnaire server 2, the questionnaire analyzer 3, and the database 4) having such functions operates as follows.

  The questionnaire server 2 requests the subject to answer the questionnaire. Note that an unspecified number of subjects may be requested in exchange for the granting of points or some kind of privilege.

  The questionnaire server 2 accepts the answers to the questionnaire, and the vocabulary of “new service that realizes a recycling-oriented society” as an evaluation target as shown in FIG. 6 and a plurality of marker targets to which the same question as the evaluation target can be applied. The vocabulary (marker objects 1 to 3 in FIG. 6) and the target information (presentation information, explanation) about each vocabulary are presented to the questionnaire input terminal 9 of each subject (corresponding to step S301 in FIG. 3).

  FIG. 6 is an example of a questionnaire presented to a subject in an adaptation example.

  In the present embodiment, the target information is presented only with characters, but photographs, videos, and sounds may be presented in addition to the characters.

  In addition, it is desirable to select a marker object that has already been embodied in order to clarify the image, but if the same questionnaire question sentence is applicable, it is not limited to belonging to the same category. .

  For example, as shown in FIG. 6, “Ubiquitous Service” (marker target 3) that is not closely related to the “recyclable society realization new service” to be evaluated may be included, or information on products rather than services may be included. That is, if the same question as the evaluation target is applicable, that is sufficient.

  The questionnaire server 2 presents the contents of the question shown in FIG. 7 to the questionnaire input terminal 9 of each subject for each subject presenting the questionnaire, and receives an answer. The answer may be received as data or may be received on another medium (corresponding to step S302 in FIG. 3).

  FIG. 7 is an example of a questionnaire response screen presented to the subject in the adaptation example.

  The data that accepts input includes score data, and further adds attribute information such as age, sex, and affiliation of each subject as necessary.

  The questionnaire server 2 totals the received score data and stores it in the database 4 (score database 20) as data for each subject and each target (corresponding to step S303 in FIG. 3).

  The questionnaire analysis device 3 obtains the score data of the marker target group of all subjects from the database 4 and analyzes the data structure of the scores of each subject, thereby digitizing the information interpretation characteristics of each subject (FIG. 3 steps S304 and S305).

  An example of the analysis method will be described. If the number of subjects is N, the number of marker targets in the marker target group is n, and the number of questions in the questionnaire is p, the data Ai of the subject i becomes n × p matrix data, and n There are N p-dimensional spaces in which marker objects are arranged.

  The data structure means a positional relationship of n objects arranged in a p-dimensional space, and by comparing the relative positional relationship of N objects, the evaluation of each subject can be evaluated. Features can be extracted. In particular, since the difference in the sensibility evaluation of the subject is small in the marker target group, most of the difference in the data structure of the score expresses the difference in interpretation of the target information presented.

  As a method of comparing the relative positional relations of objects, data such as deviation can be obtained by clustering similar subjects or comparing the average score for each question item of each subject with the score for each question item of a specific subject. Although there is a method of calculating a statistic representing the difference between the two, it is not limited to this.

  As an application example of clustering, a method of clustering all subjects into C clusters by using the value of an objective function based on orthogonal procrustis analysis as an index of similarity will be described.

  The questionnaire analysis device 3 randomly assigns N n × p matrix data, which is the matrix data of all subjects, to any one of the C clusters, and the subjects belonging to each cluster k as shown in Equation (1). By taking an average of the matrix data Ai of the set Gk, C central matrices Bk represented by an n × p matrix are calculated.

  The questionnaire analysis apparatus 3 sets an objective function F as an index indicating the difference in data structure between the matrix data Ai and the central matrix Bk of subjects belonging to each cluster, and sets each subject i to the cluster k that minimizes the objective function F. Reallocate The objective function F is calculated by calculating the distance between each Ai and the central matrix Bk rotated so as to minimize the distance from the central matrix Bk using the orthogonal transformation matrix Ti as shown in the equation (2). can get.

  Here, Lj is a p × p Lagrangian coefficient matrix, and I is a p × p unit matrix.

  Further, the questionnaire analysis apparatus 3 updates the orthogonal transformation matrix Ti using the following formula (3) and each center matrix Bk using the formula (4). It is considered that convergence has occurred when the change due to the update of the orthogonal transformation matrix Ti and each central matrix Bk is below a certain threshold, and the subject set Gk at that time is taken as the clustering result.

Note that Vi is an eigenvector of M _i M _i ^T , and Wi is an eigenvector of M _i ^T M _i , and is calculated by Expression (5).

  As another clustering method, an index representing the degree of coincidence of data structures or the closeness of distances between points is prepared, and an arbitrary threshold value or FCM (Fuzzy c-Means) method is used. You may digitize using these clustering methods.

  In general, the number of questions p in the questionnaire is 10 or more. If it is left as it is, it becomes 10 dimensions or more and becomes too complicated. Therefore, the dimensions may be reduced by using principal component analysis or the like.

  The questionnaire analysis device 3 obtains score data of all subjects including evaluation subjects of all subjects from the database 4 as score data of the evaluation subject group, and analyzes the data structure of the scores of each subject, thereby rating each subject. Are numerically converted (corresponding to steps S306 and S307 in FIG. 3).

  Note that the data structure analysis method is the same as the marker target group analysis method, but the evaluation target group includes not only the evaluation target but also the marker target. There are mixed interpretation differences.

The questionnaire analysis device 3 is based on the difference between the digitized evaluation characteristics of each subject to the marker target group and the evaluation characteristics of all the subjects evaluated by the target target group data structure analysis unit 402. Only the difference in the sensuous evaluation to the evaluation object is separated (corresponding to step S308 in FIG. 3).

  FIG. 8 is an example in which the characteristics of a subject are separated by a combination of clusters to which the subject belongs in the analysis process.

  For example, as shown in FIG. 8, the above-described separation is performed by comparing the subjects clustered by the analysis of the data structure of the marker target group with the subjects clustered by the analysis of the data structure of the evaluation target group. It is possible to identify whether or not (FIG. 8 805 to 808) is a general group or individual group and to interpret whether or not the interpretation (FIG. 8 801 to 804) is a general group or a group of special views. . In addition, other characteristics, the number of people, and the like can be extracted from the questionnaire results and various information acquired simultaneously with the questionnaire results.

  Also, using the statistics representing the difference between each subject's data by the marker target group analysis and the average value of multiple subjects by the evaluation target analysis, the predicted value of the score to each target target of each subject is estimated, and the actual value is estimated. The difference in the sensibility evaluation of each subject can be quantified by the difference from the value.

  The questionnaire analysis apparatus 3 uses information such as the classification results of the obtained subjects and the sensuous evaluation results for the evaluation target, the information shown in FIG. The data is output on the screen (corresponding to step S309 in FIG. 3).

  According to this adaptation example, a specific vocabulary can be evaluated, and an evaluation for the vocabulary can be extracted using a scaling questionnaire.

  That is, according to this adaptation example, the questionnaire system 1 can analyze the interpretation and sensitivity of the subject's vocabulary from the questionnaire of the vocabulary to be evaluated.

  According to the first embodiment of the present invention, by preparing a plurality of objects to which the same question can be applied in addition to the evaluation object, the interpretation included in the score by the data of the highly versatile scale questionnaire The difference can be analyzed, and the analysis can be applied to unknown products and services.

  In addition, according to the first embodiment of the present invention, by setting the target to be prepared in addition to the evaluation target as a marker target to which the image is fixed, it is possible to eliminate the difference in interpretation for ambiguous information presentation. It is possible to separate detailed evaluations and grasp details such as the number of people.

  Furthermore, according to the first embodiment of the present invention, since the interpretation is special but the evaluation can be used as an analysis object without excluding general subjects, data collection is expensive in proportion to the number of subjects. Cost can be reduced.

  Furthermore, according to the first embodiment of the present invention, it is necessary to prepare a plurality of targets when giving structure to data, but it is necessary to prepare a plurality of evaluation targets by including a marker target in the evaluation target group. Therefore, the cost of data collection, which is expensive in proportion to the number of questions, can be reduced.

  That is, according to the first embodiment of the present invention, the same questionnaire question is asked for a plurality of non-target objects in addition to the target evaluation object, and the data structure is added to the score data for each subject of each subject. By having it, it is possible to analyze differences in interpretation included in the score even with highly versatile scale questionnaire data.

  A second embodiment will be described with reference to FIGS.

  The second embodiment includes the same parts as those of the first embodiment. Similar parts are denoted by the same reference numerals, and detailed description thereof is omitted. Further, the second embodiment will be described by using the questionnaire system illustrated in FIG. 1 as in the first embodiment.

  FIG. 9 is a block diagram illustrating functions of the questionnaire system 1 according to the second embodiment. Compared with the questionnaire system 1 of the first embodiment, the questionnaire system 1 of the second embodiment includes a subject classification unit 303 in the interpretation difference extraction unit 30, and a marker object group data structure analysis unit 302. Based on the digitized information interpretation characteristics of each subject, the subject group to be analyzed is clustered into an arbitrary number of clusters, and the subject group belonging to an arbitrary cluster is extracted.

  In addition, the target object group score data extraction unit 401 of the sensitivity difference extraction unit 40 reads score data related to the evaluation target group of only subjects belonging to an arbitrary cluster extracted by the subject classification unit 303 from the score data database 20. Different.

  Further, the target object evaluation analysis unit 403 of the sensitivity difference extraction unit 40 calculates the difference in the sensitivity evaluation to the evaluation target from the difference in the rating characteristics of each test subject quantified by the target target group data structure analysis unit 402. It differs in the point to extract.

  FIG. 10 is a block diagram illustrating a schematic configuration of the questionnaire analysis apparatus 3 according to the second embodiment.

  The questionnaire analysis apparatus 3 according to the second embodiment is different from the questionnaire analysis apparatus 3 according to the first embodiment in that a subject classification unit 303 is provided in the interpretation difference extraction unit 30.

  The subject classification unit 303 classifies subjects into arbitrary groups based on the information interpretation characteristics of each subject, and divides them into groups such as subjects who perform general information interpretation or subjects who perform characteristic interpretation. .

  Next, the overall operation of the second exemplary embodiment of the present invention will be described.

  FIG. 11 is a flowchart showing the operation of the questionnaire system 1 according to the second embodiment.

  As in the first embodiment (see FIG. 3), the questionnaire system 1 presents a questionnaire using the score data collection unit 10 (step S301).

  The questionnaire system 1 uses the score data collection unit 10 to perform a questionnaire about the contents of the question for each subject, and transmits the result to the score data database 20 (step S302).

  The questionnaire system 1 stores the result information transmitted from the score data input unit 102 using the score data database 20 (step S303).

  The questionnaire system 1 uses the interpretation difference extraction unit 30 to extract and acquire score data related to a marker target group of a plurality of subjects from the score data database 20 (step S304).

  The questionnaire system 1 uses the interpretation difference extraction unit 30 to analyze the acquired data structure of the scores of each subject, and digitizes the interpretation characteristics of each subject's information (step S305).

  The questionnaire system 1 uses the interpretation difference extraction unit 30 to cluster each subject into an arbitrary number of clusters based on the interpretation characteristics of each subject's digitized information. Further, a subject group belonging to an arbitrary cluster is extracted from the result (step S401).

  The questionnaire system 1 is a group of a plurality of subjects including the subjects to be evaluated of a plurality of subjects who are the same as the subject whose data is extracted by the interpretation difference extraction unit 30 from the score data database 20 using the sensitivity difference extraction unit 40. The score data regarding a certain evaluation target group and the score data regarding only the test target group belonging to the arbitrary cluster extracted in the previous step are acquired (step S402).

  The questionnaire system 1 uses the sensitivity difference extraction unit 40 to analyze the data structure of the acquired scores of each subject and digitizes the rating characteristics of each subject (step S307).

  The questionnaire system 1 uses the sensitivity difference extraction unit 40 to separate the difference in the emotional evaluation of the evaluation target from the difference between the scored characteristics of the respective subjects (step S403).

  The questionnaire system 1 uses the analysis result output unit 50 to output information such as the analysis result of the evaluation on the evaluation target by each subject obtained by the sensitivity difference extraction unit 40 and the distribution of each subject (step S309).

  Through the above series of operations, the questionnaire system 1 according to the second embodiment can analyze the interpretation and sensitivity of the subject's vocabulary from the questionnaire of the vocabulary to be evaluated.

  Next, the second embodiment will be described by giving a specific example.

  In this application example, compared with the application example of the first embodiment, the questionnaire analysis apparatus 3 clusters all subjects into an arbitrary number of clusters based on the information interpretation characteristics of each subject, The difference is that the group of subjects considered to be interpreted is extracted, and the difference in the sensibility evaluation to the evaluation target is extracted by the score data on the evaluation target group of only the general subjects extracted (mainly subject classification The operation of the unit 303 is different).

  The questionnaire analysis device 3 provided with the subject classification unit 303 is given arbitrarily by using orthogonal procrutis analysis as a method of analyzing the data structure of the score data of the marker target groups of all subjects acquired from the questionnaire server 2. All subjects are clustered to an arbitrary number of clusters based on the distance of each space for each subject when the subject data of each subject is rotated so as to minimize the distance to each center point of the cluster.

  As another clustering method, there is a method of preparing an index representing the degree of coincidence of data structures and the closeness of distances between points, and classifying using an arbitrary threshold or FCM method. The method is not limited to the method described above.

  Furthermore, the questionnaire analysis apparatus 3 extracts clusters that are considered to perform general information interpretation based on the number of members of each cluster. Other methods of extracting clusters that are considered to perform general information interpretation include clusters with small differences from the average value of all subject data at the center of each cluster, and past questionnaire results of general questionnaire results. There is also a method of extracting clusters with small differences. Further, the number of clusters to be extracted is not limited to one, and there may be a plurality of clusters such as a cluster group whose difference from the comparison value is an arbitrarily set threshold value or less.

  Further, the questionnaire analysis apparatus 3 acquires, from the database 4, the score data of all subjects including the evaluation target as the score data of the evaluation target group, limited to the subject group belonging to the extracted cluster, and the score of each subject is obtained. By analyzing the data structure, the scoring characteristics of each subject are quantified.

  The numerical method is based on the method described in the application example of the first embodiment, and as a method for comparing the relative positional relationship of objects, clustering of similar subjects, the average of all subjects or the past Deviations that represent differences from the actual data are calculated.

  Furthermore, the questionnaire analysis device 3 analyzes the sensuous evaluation of the evaluation object based on the quantified rating feature.

  As an analysis method, there is a method to extract the type, characteristics, scale, etc. of the group for which special sensitivity evaluation is performed in addition to the group considered to perform general sensitivity evaluation based on the number of members in each cluster. It is done.

  In addition, it is also possible to grasp the relationship with the attributes of individual subjects using the deviation representing the difference between the average of all subjects and past performance data.

  In this example, data of subjects with different information interpretations have already been excluded, and the difference in the data structure of the scores represents only the difference in the emotional evaluation. In the analysis, unlike the first application example, it is not necessary to refer to the characteristics of the evaluation on the marker target group.

  According to this adaptation example, a specific vocabulary can be evaluated, and an evaluation for the vocabulary can be extracted using a scaling questionnaire.

  That is, according to this adaptation example, the questionnaire system 1 can analyze the interpretation and sensitivity of the subject's vocabulary from the questionnaire of the vocabulary to be evaluated.

  According to the second embodiment of the present invention, by preparing a plurality of objects to which the same question can be applied in addition to the target object, the score is also included in the highly versatile scale questionnaire data. Analysis of differences in interpretation is possible, and analysis can be applied to unknown products and services.

  In addition, according to the second embodiment of the present invention, the target to be prepared in addition to the target object is a marker object to which the image is fixed, so that it is possible to eliminate the difference in interpretation for the ambiguous information presentation. It is possible to separate detailed evaluations and grasp details such as the number of people.

  Furthermore, according to the second embodiment of the present invention, it is possible to extract only the subject group having the same interpretation of information and analyze the evaluation of the target object, thereby reducing the data processing amount.

  Furthermore, according to the second embodiment of the present invention, it is possible to reduce factors to be considered, and it is easy to understand the characteristics of each cluster.

  That is, according to the second embodiment of the present invention, by presenting information on which kansei evaluation is fixed as a marker for another evaluation target vocabulary for performing the same questionnaire question, It is possible to separate the difference in opinion and sensitivity to the evaluation object itself and the difference in interpretation and word feeling during image transmission.

  In the description of the embodiment, the target to be evaluated is described as a vocabulary, but the target to be evaluated is not limited to the vocabulary. For example, in addition to vocabulary, there are sentences, images, concepts, photographs, videos, sounds, sounds, and the like. Further, the object to be evaluated may be a combination of the exemplified ones. In other words, it is sufficient if a questionnaire can be conducted.

  Similarly, the questionnaire system 1 has been described as being divided into a plurality of devices. However, for the sake of clarity, the various devices may be arbitrarily divided.

  Similarly, the functions of various apparatuses have been described by dividing them into a plurality of parts, but for the sake of clarity, the division of the various functions is arbitrary. In other words, a plurality of devices and functions may be expressed collectively or may be expressed in a subdivided manner. For example, all the functions may be provided in one apparatus, and information acquisition and analysis / analysis processing may be performed.

  The present invention can be applied to uses such as improving the accuracy of marketing related to any product or service by classifying subjects or extracting features specialized only in sensuous evaluation.

  Furthermore, since detailed analysis of evaluations for ambiguous services and the like becomes possible, it can be applied to applications such as realization of co-creation such as open innovation.

1 is a configuration diagram illustrating a questionnaire system 1 according to an embodiment. It is a block diagram which shows the function of the questionnaire system 1 of one Embodiment. It is a flowchart which shows operation | movement of the questionnaire system 1 of one Embodiment. 3 is a block diagram illustrating a schematic configuration of a questionnaire server 2. FIG. 3 is a block diagram showing a schematic configuration of a questionnaire analysis device 3. FIG. It is an illustration of the screen of the questionnaire presented to a subject in an adaptation example. It is an illustration of the reply screen of the questionnaire presented to a subject in an adaptation example. It is an example which isolate | separates a test subject's characteristic with the combination of the cluster to which a test subject belongs in an analysis process. It is a block diagram which shows the function of the questionnaire system 1 of 2nd Embodiment. It is a block diagram which shows the schematic structure of the questionnaire analyzer 3 of one Embodiment. It is a flowchart which shows operation | movement of the questionnaire system 1 of 2nd Embodiment.

Explanation of symbols

1 Questionnaire system (questionnaire analysis system)
2 Questionnaire Server 3 Questionnaire Analyzer 4 Database 8 Network 9 Questionnaire Input Terminal 10 Rating Data Collection Unit (Questionnaire and Questionnaire Result Acquisition Unit)
20 Rating data database (storage)
30 Interpretation Difference Extraction Unit 40 Sensitivity Difference Extraction Unit 50 Analysis Result Output Unit 101 Target Information Presentation Unit 102 Rating Data Input Unit (Questionnaire Result Information Input Unit)
301 Marker target group score data extraction unit 302 Marker target group score data extraction unit 303 Subject classification unit 401 Target target group score data extraction unit 402 Target target group data structure analysis unit 403 Target target evaluation analysis unit

Claims (17)

In a questionnaire analysis system that conducts questionnaires on multiple subjects for the subjects to be evaluated, and summarizes and analyzes the results of the questionnaires,
The questionnaire analysis system
A control unit that performs various types of information processing;
A questionnaire result information input unit for receiving answers to a questionnaire regarding a target to be evaluated and a marker with a small difference in interpretation between the target to be evaluated presented to multiple subjects;
A storage unit for storing and storing various types of information including information on answers to the questionnaire acquired by the questionnaire result information input unit;
Comprising at least
The controller is
When carrying out the questionnaire, along with presents and the subject of interest and a plurality of the marker to be evaluated in a plurality of subjects, line the question of similar content to the target that the target you want to the evaluation serving as the plurality of markers and scores the data collection means scores survey results of a plurality of people to get through the questionnaire result information input section, and records in the storage unit extracts processed scores data from the rating questionnaire results obtained I,
Scoring data about the target to be more markers that are recorded from the storage unit and acquires as a marker control group for each subject, relative to other subjects relating to the interpretation of each subject from the feature scores data of the marker object group Interpretation difference extraction means for extracting and processing features in which a general relationship appears ,
The score data related to the target group including the target to be evaluated recorded from the storage unit is obtained for each subject, and there is a relative relationship with other subjects regarding how to assign the score for each subject from the characteristics of the obtained score data. is extracted processes appeared characteristic, relative characteristics for other subjects on extraction predetermined about how scores subjects to other subjects of the relative characteristics and the interpretation and extracted with difference extracting means appropriate subject interpretation comparing the difference, the sensitivity difference extracting means for separating the difference in relative sensitivity assessment for other subjects to the subject to be the evaluation with the result of the comparison of the predetermined subject,
Including
A questionnaire analysis system characterized in that the interpretation and sensitivity of a predetermined subject reflected in a score questionnaire result including sensitivity and interpretation are combined so as to be relatively comparable between subjects . The interpretation difference extraction means includes
The score data regarding the marker target group of a plurality of subjects recorded from the storage unit is read, and the interpretation characteristics for each of the plurality of subjects are determined from the differences between the subjects in the data structure of the score data of the read subjects. Extract and process
The sensitivity difference extraction means includes:
Wherein each of the plurality of subjects from the difference of the data structure of the score data of the plurality of subjects the acquired acquires score data relating to the target group for each subject including the evaluation of the plurality of subjects that have been recorded from the storage unit along with extracting process the characteristics of grading classification differences in sensibility evaluation of the difference between the grading features of each of the plurality of features and the plurality of subjects interpretations for each subject to be evaluated for each subject The questionnaire analysis system according to claim 1, wherein: The process of extracting the features of the interpretation of the plurality of subjects clustering the plurality of subjects into an arbitrary subject group from the difference in the data structure of the score data of the plurality of subjects obtained from the storage unit,
The process of extracting the scoring characteristics of the plurality of subjects is performed by clustering the subject groups of any of the plurality of subjects from the difference in the data structure of the score data of the plurality of subjects obtained from the storage unit,
Questionnaire analysis system according to claim 1 or 2 Kansei evaluation from both the result of differences of clustering to be evaluated is characterized <br/> separating the different groups of subjects. The process of extracting the interpretation features of the plurality of subjects is
Calculate a statistic by comparing the average score for each question item of the plurality of subjects and the score for each question item of a specific subject,
The process of extracting the scoring characteristics of the plurality of subjects is
A statistic representing a difference between the data of the plurality of subjects, and an estimated value of the score to the evaluation target of the plurality of subjects from an average value of the plurality of subjects by the evaluation target analysis,
The questionnaire analysis system according to claim 1 or 2, wherein a difference in sensibility evaluation of a plurality of subjects is extracted based on a difference from an actual value. The interpretation difference extraction means includes
And extracted acquiring and processing scores data regarding marker object group of a plurality of subjects from the storage unit, using the feature of the interpretation of each subject, and clustered into any cluster based on the feature of interpreting the plurality of subjects, said Extract a group of subjects belonging to any cluster,
The sensitivity difference extraction means includes:
Using the feature of the grading of the from the storage unit the plurality of extracted acquiring and processing scores data related to the evaluation objects set of subjects belonging to the arbitrary cluster subject of grading features of each of the plurality of subjects questionnaire analysis system according to claim 1, wherein <br/> be extracted as a difference in relative sensitivity assessment for other subjects for each subject the difference to the interpretation of the differences excluded were evaluated. A questionnaire analysis system according to any one of claims 1 to 5 ,
The questionnaire analysis system
Including a questionnaire server and a questionnaire analyzer,
The control part of the questionnaire server includes the score data collecting means ,
The control unit of the questionnaire analyzer includes interpretation difference extraction means and sensitivity difference extraction means .
A questionnaire analysis system characterized by comprising . A questionnaire server included in the questionnaire analysis system according to claim 6 . A questionnaire analysis device included in the questionnaire analysis system according to claim 6 . A questionnaire analysis method using a computer system that conducts a questionnaire on a plurality of subjects for a subject to be evaluated, totals the results of the questionnaire, and totals and analyzes the results of the questionnaire,
The computer system includes obtaining means for obtaining the questionnaire answers for subjects difference interpretation by between objects and subjects to be evaluated were presented in a plurality of persons subject to a small marker accepted, as the questionnaire result information, the questionnaire result information And having means for recording the information on the questionnaire response obtained by the means to the storage unit ,
The computer system is
When carrying out the questionnaire, along with the subject you want to evaluate, together with presenting the subject to be a multiple of the marker in a plurality of subjects, line the question of similar content to the target that the target you want to the evaluation serving as the plurality of markers and scores the data collection means scores survey results of a plurality of people to get through the means for acquiring the questionnaire result information, and records in the storage unit extracts processed scores data from the rating questionnaire results obtained I,
Scoring data about the target to be more markers that are recorded from the storage unit and acquires as a marker control group for each subject, relative to other subjects relating to the interpretation of each subject from the feature scores data of the marker object group Interpretation difference extraction means for extracting and processing features in which a general relationship appears ,
Wherein the relative relationship with other subjects relating Tsukekata scores for each subject from the feature scores data to obtain the score data the acquired related to the target group appeared including the target to be evaluated are recorded from the storage unit is extracted processes, the difference between the relative characteristics to other subjects of the relative characteristics and the interpretation and extracted with difference extracting means corresponding subject interpretation of other subjects on extraction predetermined about how scores subjects comparison, and sensitivity difference extracting means for separating the difference in relative sensitivity assessment for other subjects to the subject to be the evaluation with the result of the comparison of the predetermined subject,
Including
Questionnaire analysis, characterized by the fact that the interpretation and sensitivity of a given subject reflected in the result of the questionnaire survey that includes both sensitivity and interpretation are classified so as to be relatively comparable between subjects. Method. The interpretation difference extraction means included in the computer system includes:
The score data regarding the marker target group of a plurality of subjects recorded from the storage unit is read, and the interpretation characteristics for each of the plurality of subjects are determined from the differences between the subjects in the data structure of the score data of the read subjects. Extract and process
The sensitivity difference extraction means includes:
Wherein each of the plurality of subjects from the difference of the data structure of the score data of the plurality of subjects the acquired acquires score data relating to the target group for each subject including the evaluation of the plurality of subjects that have been recorded from the storage unit along with extracting process the characteristics of grading classification differences in sensibility evaluation of the difference between the grading features of each of the plurality of features and the plurality of subjects interpretations for each subject to be evaluated for each subject The questionnaire analysis method according to claim 9, wherein: The process of extracting the interpretation features of the plurality of subjects recognizes the difference in the data structure of the score data of the plurality of subjects acquired from the storage unit, and clusters the plurality of subjects into an arbitrary subject group,
The process of extracting the scoring characteristics of the plurality of subjects is
Recognizing the difference in data structure of the score data of a plurality of subjects acquired from the storage unit, clustering the plurality of subjects into an arbitrary subject group,
The questionnaire analysis method according to claim 9 or 10, wherein a difference between the two clustered results is recognized, and a group of subjects having different sensibility evaluations on evaluation targets is separated. In the process of extracting the interpretation features of the plurality of subjects,
Calculate a statistic by comparing the average score for each question item of the plurality of subjects and the score for each question item of a specific subject,
In the process of extracting the scoring characteristics of the plurality of subjects,
The statistical value representing the difference between the data of the plurality of subjects and the predicted value of the score to the evaluation target of the plurality of subjects from the average value of the plurality of subjects by the evaluation target analysis are estimated, and a plurality of values are calculated depending on the difference from the actual value. The questionnaire analysis method according to claim 9 or 10, wherein a difference in sensibility evaluation of the subjects is extracted. The interpretation difference extraction means includes
Using the feature of the interpretation of each subject extracted acquiring and processing scores data regarding marker object group of a plurality of subjects from the storage unit, and clustered into any cluster based on the feature of interpreting the plurality of subjects, said optional Extract a group of subjects belonging to the cluster of
The sensitivity difference extraction means includes:
Using the feature of the grading of the from the storage unit the plurality of extracted acquiring and processing scores data related to the evaluation objects set of subjects belonging to the arbitrary cluster subject of grading features of each of the plurality of subjects The questionnaire analysis method according to claim 9, wherein the difference is extracted as a difference in relative sensibility evaluation with respect to another subject for each subject to an evaluation target from which a difference in interpretation is removed . A questionnaire analysis method according to any one of claims 9 to 13 ,
The computer system includes:
Including a questionnaire server for conducting a questionnaire, and a questionnaire analyzer for analyzing and analyzing the questionnaire results ,
The control part of the questionnaire server includes the score data collecting means,
Wherein that the control unit of the questionnaire analysis system comprises comprises the sensibility difference extracting means and the interpretation difference extracting means, questionnaire analysis method is realized. In a program that operates a questionnaire analysis system that conducts a questionnaire on the target for multiple subjects and aggregates and analyzes the results of the questionnaire,
The questionnaire analysis system has at least a control unit that performs various information processing, and questionnaire result information input that accepts a questionnaire response regarding a target that is a marker with a small difference in interpretation between the target to be evaluated and the target presented to a plurality of subjects. And a storage unit for storing and storing various types of information including information on the questionnaire responses acquired by the questionnaire result information input unit,
The program controls the control unit,
When carrying out the questionnaire, along with presents and the subject of interest and a plurality of the marker to be evaluated in a plurality of subjects, line the question of similar content to the target that the target you want to the evaluation serving as the plurality of markers and scores the data collection means scores survey results of a plurality of people to get through the questionnaire result information input section, and records in the storage unit extracts processed scores data from the rating questionnaire results obtained I,
Scoring data about the target to be more markers that are recorded from the storage unit and acquires as a marker control group for each subject, relative to other subjects relating to the interpretation of each subject from the feature scores data of the marker object group Interpretation difference extraction means for extracting and processing features in which a general relationship appears ,
The score data related to the target group including the target to be evaluated recorded from the storage unit is obtained for each subject, and there is a relative relationship with other subjects regarding how to assign the score for each subject from the characteristics of the obtained score data. is extracted processes appeared characteristic, relative characteristics for other subjects on extraction predetermined about how scores subjects to other subjects of the relative characteristics and the interpretation and extracted with difference extracting means appropriate subject interpretation Comparing the difference of, and functioning as a sensitivity difference extraction means for separating the difference of the relative sensitivity evaluation to other subjects to the subject to be evaluated that the predetermined subject has from the comparison result ,
A program characterized in that the interpretation and sensitivity of a predetermined subject reflected in a score questionnaire result including sensitivity and interpretation are combined so that the subjects can be relatively compared . A program constituting a part of a program for causing the questionnaire analysis system according to claim 15 to function,
A questionnaire server for conducting a questionnaire that forms part of the questionnaire analysis system;
A score data collection program that causes a control unit of the questionnaire server to function as the score data collection means. A program constituting a part of a program for causing the questionnaire analysis system according to claim 15 to function,
A questionnaire analysis device for analyzing a questionnaire result, which is a part of the questionnaire analysis system, functions.
A rating data analysis program which causes the control unit of the questionnaire analysis apparatus to function as the interpretation difference extraction means and the sensitivity difference extraction means.

Images