JP6813917B1 - Questionnaire result analysis system and questionnaire result analysis method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a questionnaire result analysis system capable of grasping a new population capable of obtaining reaction data highly useful for marketing. An analysis system S has a tendency value recognition unit 1j that recognizes a tendency value of a population and a tendency value of a response of a target group extracted from the population, and a degree of deviation that recognizes the degree of deviation of those tendency values. It includes a recognition unit 1k and a peculiar keyword / tag recognition unit 1l that recognizes a keyword related to a questionnaire as a peculiar keyword of a target group when the degree of deviation is equal to or higher than a predetermined threshold value. [Selection diagram] Fig. 2

Description

The present invention is a questionnaire result analysis system for analyzing a population consisting of a plurality of respondents who answered the questions based on the answers to the questions constituting the questionnaire, or the questionnaire itself, and a questionnaire result analysis. Regarding the method.

In recent years, a method of acquiring a large amount of direct reaction data from an individual and using the reaction data for marketing has been proposed. Reaction data that can be used in this way includes, for example, answers to questionnaires.

When adopting answers to a questionnaire as reaction data, the reliability of the questionnaire (and thus its usefulness in marketing) depends on how the population of multiple respondents to be conducted is determined. ) Will change.

Therefore, there is known a method of analyzing the characteristics of the entire population by determining a sample population based on the characteristics of each respondent included in the population and analyzing the sample population (for example, a patent). Reference 1).

Japanese Patent No. 5334181

By the way, when conducting marketing, it is important to properly grasp the characteristics of the population that are predicted to be targets for sales, advertisements, etc., but on the other hand, it has characteristics that are useful for marketing. It is also important to discover a new population (for example, a population that is likely to be purchased).

However, the method described in Patent Document 1 is a method for appropriately grasping the characteristics of the entire population, which is predicted to be a target for sales, advertisements, etc. in advance. Therefore, in the method described in Patent Document 1, it is possible to grasp whether the population has peculiarity in marketing in other fields (and by extension, is the population capable of obtaining highly useful reaction data). I couldn't.

The present invention has been made in view of the above points, and provides a questionnaire result analysis system capable of grasping a new population capable of obtaining reaction data highly useful for marketing, and a questionnaire result analysis method. The purpose is.

The questionnaire result analysis system of the present invention
It is a questionnaire result analysis system for analyzing a population consisting of a plurality of respondents who answered the above questions based on the answers to the questions constituting the questionnaire.
A second questionnaire consisting of a plurality of questions containing at least one question different from the plurality of questions constituting the first questionnaire is based on the questions included in the second questionnaire. , A process of recognizing a keyword included in the question as a keyword related to the second questionnaire, and a process of assigning a tag indicating the attribute of the question to the second questionnaire as a tag related to the second questionnaire. A second keyword / tag processing unit that executes at least one,
The first answer recognition unit that recognizes the answer by the population to the first questionnaire, and
The second answer recognition unit that recognizes the answer by the population to the second questionnaire, and
Recognizing target group recognition unit as the target group the group that was given the answer against a predetermined questions contained in the first survey of the population,
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, and a value based on the tendency of the response of the target group to the second questionnaire. The group tendency value recognition unit that recognizes as the group tendency value that is
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population, and a value based on the tendency of the population to answer the second questionnaire. The population inclination recognition unit that recognizes as the population tendency value , which is
A divergence degree recognition unit that recognizes the degree of divergence, which is a value indicating the degree of divergence between the group tendency value and the population tendency value based on the group tendency value and the population tendency value.
When the degree of divergence is equal to or greater than a predetermined threshold value, at least one of the keyword and the tag related to the second questionnaire is a specific keyword or a tag that is specific to the target group. Alternatively, it is characterized by having a peculiar keyword / tag recognition unit that is recognized as a peculiar tag.

As described above, the questionnaire result analysis system of the present invention recognizes the keywords included in the questions constituting the second questionnaire as the keywords related to the second questionnaire, and the question. At least one of the processes of assigning a tag indicating the attribute of to the second questionnaire as a tag related to the second questionnaire is executed. That is, in this analysis system, keywords or tags (and thus, the field of marketing) that can be examined from the answers to the second questionnaire are set.

After that, the system recognizes the target group from the population based on the response to the first questionnaire, and based on the response to the second questionnaire, the tendency value of the target group and the tendency value of the population Recognize the degree of divergence. Further, after that, the system recognizes whether or not the keyword or tag related to the second questionnaire has specificity for the target group based on the degree of deviation.

Here, the population for which the tendency value is recognized may be a group that can grasp a rough tendency of the entire population. Therefore, this population includes not only the entire population including the target group, but also all the respondents other than the target group included in the population.

Thereby, according to this system, it is possible to recognize a group having peculiarity in the field of marketing regarding a peculiar keyword or a peculiar tag from the population. Then, for example, when the population is determined by adopting or excluding the group, reaction data highly useful in the field of marketing can be obtained from the population.

Further, in the questionnaire result analysis system of the present invention,
A third questionnaire composed of a plurality of questions constituting the first questionnaire and a plurality of questions including at least one question different from the plurality of questions constituting the second questionnaire. On the other hand, based on the question included in the third questionnaire, the process of recognizing the keyword contained in the question as a keyword related to the third questionnaire, and the tag indicating the attribute of the question are attached. A third keyword / tag processing unit that executes at least one of the processes assigned to the third questionnaire as a tag related to the third questionnaire, and
A matching degree recognition unit that recognizes the degree of matching between at least one of the peculiar keyword and the peculiar tag and at least one of the keyword and the tag related to the third questionnaire.
When the degree of agreement is equal to or higher than a predetermined threshold value, the target group is proposed as a population to carry out the third questionnaire, or is proposed as a group to be excluded from the population to be carried out. It is preferable to have.

If the keyword or tag for the third survey matches a singular keyword or tag that has specificity for a group, then the keyword or tag for the third survey also has specificity for that group. There is a high probability that you are doing it. That is, the group is likely to have a peculiar tendency also in the field of marketing related to the third questionnaire.

Therefore, if the group recognized based on at least one of the singular keyword and the singular tag is proposed as the population for conducting the third questionnaire, or as a group to be excluded from the population for conducting the third questionnaire, the first There is a high probability that a specific answer to the third questionnaire (that is, reaction data that is highly useful in the field of marketing related to the third questionnaire) can be obtained.

Moreover, the questionnaire result analysis method of this invention
It is a questionnaire result analysis method for analyzing a population consisting of a plurality of respondents who answered the above based on the answers to the questions constituting the questionnaire.
In contrast to the second questionnaire, in which the second keyword / tag processing unit is composed of a plurality of questions including at least one question different from the plurality of questions constituting the first questionnaire, the second questionnaire is described. Based on the question contained in, the process of recognizing the keyword contained in the question as a keyword related to the second questionnaire, and the tag indicating the attribute of the question as the tag related to the second questionnaire. The process of executing at least one of the processes given to the second questionnaire, and
The process in which the first answer recognition unit recognizes the answer by the population to the first questionnaire,
The process in which the second answer recognition unit recognizes the answer by the population to the second questionnaire,
Target group recognition unit, and a step of recognizing a target group of the group performing a predetermined answer by pairs in a predetermined questions contained in the first survey of the population,
The group tendency value recognition unit sets a value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, to the target group for the second questionnaire. The process of recognizing as a group tendency value , which is a value based on the tendency of the answer of
The value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population by the population inclination recognition unit with respect to the second questionnaire. The process of recognizing as a population tendency value , which is a value based on the population's response tendency,
A step of recognizing the degree of dissociation, which is a value indicating the degree of dissociation between the group tendency value and the population tendency value, based on the group tendency value and the population tendency value.
When the degree of deviation is equal to or greater than a predetermined threshold value, the singular keyword / tag recognition unit has specificity for the target group for at least one of the keyword and the tag related to the second questionnaire. It is characterized by including a step of recognizing it as a peculiar keyword or a peculiar tag which is a keyword or tag.

Further, the questionnaire result analysis system of the present invention is
A questionnaire result analysis system for analyzing the questionnaire based on the answers to the questions that make up the questionnaire.
For a first questionnaire composed of a plurality of questions, a process of recognizing a keyword included in the question as a keyword related to the first questionnaire based on the question included in the first questionnaire. , And a first keyword / tag processing unit that executes at least one of the processes of assigning a tag indicating the attribute of the question to the first questionnaire as a tag related to the first questionnaire.
Based on the questions included in the second questionnaire, as opposed to the second questionnaire composed of a plurality of questions containing at least one question different from the plurality of questions constituting the first questionnaire. The process of recognizing the keyword included in the question as a keyword related to the second questionnaire, and the process of assigning a tag indicating the attribute of the question to the second questionnaire as a tag related to the second questionnaire. A second keyword / tag processing unit that executes at least one of
The first answer recognition unit that recognizes the answers by the population to the first questionnaire, and
The second answer recognition unit that recognizes the answer by the population to the second questionnaire, and
Recognizing target group recognition unit as the target group the group that was given the answer against a predetermined questions contained in the first survey of the population,
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, and a value based on the tendency of the response of the target group to the second questionnaire. The group tendency value recognition unit that recognizes as the group tendency value that is
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population, and a value based on the tendency of the population to answer the second questionnaire. The population inclination recognition unit that recognizes as the population tendency value , which is
A divergence degree recognition unit that recognizes the degree of divergence, which is a value indicating the degree of divergence between the group tendency value and the population tendency value based on the group tendency value and the population tendency value.
When the degree of divergence is equal to or greater than a predetermined threshold value, the specificity of at least one of the keyword and the tag related to the second questionnaire is given to at least one of the keyword and the tag related to the first questionnaire. It is characterized by including a peculiar keyword / tag recognition unit that recognizes a peculiar keyword or a peculiar tag that is a possessed keyword or tag.

As described above, the questionnaire result analysis system of the present invention recognizes the keywords included in the questions constituting the questionnaire as the keywords related to the questionnaire for the first questionnaire and the second questionnaire, and , At least one of the processes of assigning a tag indicating the attribute of the question to the questionnaire as a tag related to the questionnaire is executed. That is, in this system, keywords or tags (and thus, the field of marketing) that can be examined from the answers of the first questionnaire and the second questionnaire are set.

After that, the system recognizes the target group from the population based on the response to the first questionnaire, and the tendency value of the target group and the tendency value of the entire population based on the response to the second questionnaire. Recognize the degree of divergence from. Further, after that, the system recognizes whether or not the keyword or tag related to the second questionnaire has specificity to the keyword or tag related to the first questionnaire based on the degree of deviation.

Here, the population for which the tendency value is recognized may be a group that can grasp a rough tendency of the entire population. Therefore, this population includes not only the entire population including the target group, but also all the respondents other than the target group included in the population.

As a result, according to this system, from the keywords or tags related to the marketing field related to the first questionnaire, the keywords or tags for extracting the group having specificity in the marketing field related to the peculiar keyword or the peculiar tag. Can recognize tags. Then, when the population is determined based on the keyword or tag, reaction data highly useful in the field of marketing can be obtained from the population.

Further, in the questionnaire result analysis system of the present invention,
A third questionnaire composed of a plurality of questions constituting the first questionnaire and a plurality of questions including at least one question different from the plurality of questions constituting the second questionnaire. On the other hand, based on the question constituting the third questionnaire, the process of recognizing the keyword included in the question as the keyword related to the third questionnaire, and the tag indicating the attribute of the question are attached. A third keyword / tag processing unit that executes at least one of the processes assigned to the third questionnaire as a tag related to the third questionnaire, and
A matching degree recognition unit that recognizes the degree of matching between at least one of the peculiar keyword and the peculiar tag and at least one of the keyword and the tag related to the third questionnaire.
When the degree of agreement is equal to or higher than a predetermined threshold value, at least one of the peculiar keyword and the peculiar tag is proposed as a keyword or tag to be considered when determining a population for conducting the third questionnaire. It is preferable to have a keyword / tag proposal section.

If the keyword or tag related to the third questionnaire matches the singular keyword or singular tag, the group corresponding to the singular keyword or singular tag has specificity to the keyword or tag related to the third questionnaire. There is a high probability that you are doing it. That is, the group is likely to have a peculiar tendency also in the field of marketing related to the third questionnaire.

Therefore, if at least one of the peculiar keyword and the peculiar tag is proposed as a keyword or tag to be considered when determining the population to be subjected to the third questionnaire in this way, it is peculiar to the third questionnaire. There is a high probability that an answer (that is, reaction data that is highly useful in the field of marketing regarding the third questionnaire) can be obtained.

Moreover, the questionnaire result analysis method of this invention
A questionnaire result analysis method for analyzing the questionnaire based on the answers to the questions that make up the questionnaire.
For the first questionnaire composed of a plurality of questions, the first keyword / tag processing unit sets the keywords included in the question based on the questions included in the first questionnaire. A step of executing at least one of a process of recognizing as a keyword related to the questionnaire 1 and a process of assigning a tag indicating the attribute of the question to the first questionnaire as a tag related to the first questionnaire.
For the second questionnaire, the second keyword / tag processing unit is composed of a plurality of questions including at least one question different from the plurality of questions constituting the first questionnaire. Based on the question included in the questionnaire, the process of recognizing the keyword contained in the question as the keyword related to the second questionnaire, and the tag indicating the attribute of the question as the tag related to the second questionnaire. A step of executing at least one of the processes given to the second questionnaire, and
The process in which the first answer recognition unit recognizes the answers by the population to the first questionnaire,
The process in which the second answer recognition unit recognizes the answer by the population to the second questionnaire,
Target group recognition unit, and a step of recognizing a target group of the group performing a predetermined answer by pairs in a predetermined questions contained in the first survey of the population,
The group tendency value recognition unit sets a value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, to the target group for the second questionnaire. The process of recognizing as a group tendency value , which is a value based on the tendency of the answer of
The value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population by the population inclination recognition unit with respect to the second questionnaire. The process of recognizing as a population tendency value , which is a value based on the population's response tendency,
A step of recognizing the degree of dissociation, which is a value indicating the degree of dissociation between the group tendency value and the population tendency value, based on the group tendency value and the population tendency value.
When the degree of divergence is equal to or greater than a predetermined threshold value, the singular keyword / tag recognition unit can use at least one of the keyword and the tag related to the second questionnaire to the keyword and the tag related to the first questionnaire. It is characterized by including a step of recognizing it as a peculiar keyword or a peculiar tag which is a keyword or tag having peculiarity to at least one of them.

Explanatory drawing which shows the schematic structure of the analysis system which concerns on embodiment. The block diagram which shows the structure of the processing part of the analysis system of FIG. The flowchart which shows the process which the analysis system of FIG. 1 executes when performing the first questionnaire from a client. FIG. 5 is a flowchart showing a process executed when the analysis system of FIG. 1 conducts a questionnaire to a customer. FIG. 5 is a flowchart showing a process to be executed when the analysis system of FIG. 1 receives a request for a second and subsequent questionnaires from a client.

Hereinafter, the configuration of the analysis system S (questionnaire result analysis system) according to the embodiment will be described with reference to the drawings.

In the following description, a manufacturer or the like who wishes to conduct a questionnaire for marketing using the analysis system S is referred to as a “client”. Respondents who answer the questionnaire sent from the analysis system S are called "customers".

The analysis system S is a system that analyzes a population consisting of a plurality of customers who have answered and the questionnaire itself based on the answers to the questions constituting the questionnaire to be conducted.

Then, the analysis system S provides the client with a service such as conducting a questionnaire according to the request from the client based on the result of the analysis and proposing a population for conducting the questionnaire. On the other hand, in this service, a predetermined reward (for example, points that can be used when purchasing a client's product) may be paid to a customer when answering a questionnaire.

[Outline configuration of system]
First, the schematic configuration of the analysis system S will be described with reference to FIG.

As shown in FIG. 1, the analysis system S is composed of a server 1 owned by a service provider by the analysis system S. The server 1 is configured to enable mutual information communication with a client terminal 2 such as a personal computer or tablet owned by the client and a customer terminal 3 such as a smartphone or tablet owned by the customer via an internet line or the like. ..

The present invention is not limited to such a configuration, and may be configured so that the system can recognize the request from the client and the response from the customer.

Therefore, for example, instead of sending the answer input via the customer terminal to the server, the answer may be separately converted into data and the data may be read directly into the server. Further, for example, the entire system may be configured by a plurality of servers. Further, at least one of the functions of the processing unit described later may be mounted on the client terminal and the customer terminal, and the system may be configured by using them.

[Configuration of each processing unit]
As shown in FIG. 2, the server 1 has a questionnaire determination unit 1a, a keyword tag storage unit 1b, and a keyword tag processing unit 1c as functions (processing units) realized by the implemented hardware configuration or program. , Customer data storage unit 1d, proposal unit 1e, population determination unit 1f, questionnaire execution unit 1g, answer recognition unit 1h, target group recognition unit 1i, tendency value recognition unit 1j, and deviation degree recognition. A unit 1k, a peculiar keyword / tag recognition unit 1l, and a matching degree recognition unit 1m are provided.

The questionnaire determination unit 1a determines a questionnaire consisting of a plurality of questions in response to a request from the client. Specifically, the client recognizes the information that the client wants to know (information that the client wants to know, etc.), and creates a plurality of questions based on the information and the questionnaire, etc. conducted in the past.

Here, the information that is the object of the questionnaire is recognized based on, for example, an answer to a predetermined template presented to the client in advance via the client terminal 2.

As the questions that make up the questionnaire, for example, a state question (for example, a question based on the mind flow) for confirming the customer's impression of the product to be asked is a two-choice answer, a multi-choice answer, and a five-step evaluation. Questions such as answers and NPS evaluation answers can be mentioned.

The keyword tag storage unit 1b stores keyword tags that may be related to the questionnaire to be conducted. Specifically, it stores data showing the relationship between a predetermined keyword and a tag set for the keyword from the viewpoint of 4P (product, sales channel, price, promotion) analysis and the like.

Further, the keyword / tag storage unit 1b stores the peculiar keyword and the peculiar tag described later in association with the stored keyword or tag.

The keyword / tag processing unit 1c (first keyword / tag processing unit, second keyword / tag processing unit, third keyword / tag processing unit) constitutes the questionnaire for the questionnaire determined by the questionnaire determination unit 1a. Based on the question you are asking, recognize the keywords contained in that question as keywords related to the survey.

Further, the keyword / tag processing unit 1c refers to the data stored in the keyword / tag storage unit 1b for the questionnaire, and sets a tag indicating the attribute of the question as a tag related to the questionnaire in the questionnaire. Give.

Examples of the process of recognizing a keyword include a process of executing text mining on a question sentence and recognizing a noun that is the subject of the sentence as a keyword based on the result.

Examples of the keywords include category keywords such as "automobile" and "travel" indicating a product or the like related to a question, a brand keyword indicating the brand of the product, a product name keyword indicating a product name, and the like.

Further, as the process of adding a tag, for example, by referring to the data stored in the keyword / tag storage unit 1b, a word indicating a higher-level concept of the recognized keyword, a word generally recalled, or the like can be used. Examples include the process of assigning a tag for each question or a tag for the entire questionnaire.

As tags, for example, brands such as "electric appliances" and "food", words that are a superordinate concept of keywords that refer to products, brands such as "convenience store" and "drug store", and sales channels related to keywords that refer to products are recalled. Words that are recalled for keywords that refer to prices such as "low price" and "luxury goods", and as a promotion route for keywords that refer to brands and products such as "TV CM" and "magazines". Words and the like can be mentioned.

In the analysis system S, the keyword / tag processing unit related to the first questionnaire and the second questionnaire composed of a plurality of questions including at least one question different from the questions constituting the first questionnaire. A third, which is composed of a keyword / tag processing unit related to, a plurality of questions including at least one question that constitutes the first questionnaire, and a question that is different from the question that constitutes the second questionnaire. The keyword / tag processing unit related to the questionnaire is configured as one processing unit.

However, the present invention is not limited to such a configuration, and any of those keyword / tag processing units may be configured as independent ones.

The customer data storage unit 1d stores information about customers who can be the target of the questionnaire. Specifically, a plurality of customers registered in the service by the analysis system S are categorized for each attribute and stored in a state in which an individual cannot be identified.

Customer attributes include, for example, age, gender, place of residence, hobbies, family structure, terminal information used for answers, questionnaires answered so far and answers to them, and specific keywords and tags described later. Can be mentioned.

The proposal unit 1e (population proposal unit, keyword / tag proposal unit) provides the client with a population capable of conducting a questionnaire and the implementation of the questionnaire (that is, including it in the population) via the client terminal 2. Suggest groups that are recommended or not recommended, and at least one of the keywords and tags that should be considered when determining the population to be surveyed.

Specifically, the proposal unit 1e searches for information about the customer from the customer data storage unit 1d based on at least one of the keyword recognized and the attached tag for the questionnaire determined by the questionnaire determination unit 1a. Then, the population that can carry out the questionnaire is presented.

In addition, the Proposal Department 1e presents the population, groups that specifically recommend the implementation of the questionnaire, groups that do not recommend (exclude) the implementation of the questionnaire, and keywords and reasons for the questionnaire. Present the tag. The content of this presentation is determined by referring to the peculiar keywords and peculiar tags described later and the degree of agreement.

In addition, the proposal unit 1e corresponds to at least one of the keywords and tags to be considered when determining the population to be surveyed from the population together with the presentation of the population, and the keywords or tags. Present a group of customers and the attributes of that group. The content of this presentation is determined by referring to the peculiar keywords and peculiar tags described later and the degree of agreement.

In the analysis system S, a population proposal unit that makes proposals about the population and a keyword / tag proposal unit that makes proposals about keywords and tags are configured as one processing unit. However, the present invention is not limited to such a configuration, and the population proposal unit and the keyword / tag proposal unit may be configured as processing units that are independent of each other.

Further, in the analysis system S, a population proposal unit that makes proposals about the population and a keyword / tag proposal unit that makes proposals about keywords and tags in order to present materials for consideration when the client determines the population. Is equipped with. However, the present invention is not limited to such a configuration, and at least one of the population proposal unit and the keyword / tag proposal unit may be omitted.

The population determination unit 1f determines the population to be surveyed from the customer data stored in the customer data storage unit 1d based on the input from the client to the client terminal 2.

The questionnaire implementation unit 1g transmits the questionnaire determined by the questionnaire determination unit 1a to the customer terminal 3 of the customer included in the population determined by the population determination unit 1f in a format in which an answer can be input. To carry out the questionnaire.

The answer recognition unit 1h (first answer recognition unit, second answer recognition unit) recognizes the questionnaire conducted by the questionnaire implementation unit 1g, and responds to the questionnaire based on the input from the customer to the customer terminal 3. recognize. In addition, the answer recognition unit 1h presents the conducted questionnaire and the answers in a predetermined format to the client via the client terminal 2.

The analysis system S recognizes the answers to the questionnaires for each customer by distinguishing the customer terminals 3. However, the present invention is not limited to such a configuration, and may be configured so that the correspondence between the questionnaire and the respondents who responded to the questionnaire can be recognized.

For example, if the respondent is configured to input the ID assigned to each respondent in advance when giving an answer, the analysis system S does not have to distinguish the terminals. Then, in that case, the respondent may use a different terminal for each questionnaire.

In the analysis system S, the first answer recognition unit that recognizes the response by the population to the first questionnaire and the second answer recognition unit that recognizes the answer by the population to the second questionnaire are used as one processing unit. It is configured. However, the present invention is not limited to such a configuration, and the first response recognition unit and the second response recognition unit may be configured as processing units independent of each other.

The target group recognition unit 1i verifies the peculiar keywords and peculiar tags described later based on the answers to the questionnaire from the population in which the predetermined questionnaire was conducted based on the answers recognized by the answer recognition unit 1h. Recognize the target group for.

The rule for recognizing the target group may be appropriately set by the system designer, may be artificially set by the system designer himself, or may be set based on the result of machine learning using AI. You may.

As a rule for recognizing a target group, there is a method in which a group of respondents who have given a predetermined answer to a predetermined question is set as the target group. The predetermined question is, for example, a question showing a policy for the entire questionnaire. In addition, specifically, regarding the product that is the theme of the questionnaire, there are questions asking about the contents such as cognition, interest, liking, behavior, comparison, purchasing, possession, use, and attachment according to the mind flow.

Questions based on the mind flow include, for example, comparing whether you know or do not know the product that is the subject of the questionnaire, whether you are interested or not, whether you like it or not, whether you have researched it, etc. Whether you have purchased it, whether you own it (or have it), whether you are using it (whether you were using it), whether you are attached to it if you are using it, etc. Questions etc. can be raised.

In the analysis system S, as the questions constituting the questionnaire, a two-choice format of "yes" and "no", or a three-choice format in which "hold" is added to them is adopted. However, the present invention is not limited to such a configuration, and a questionnaire may be constructed using questions in the form of a multi-step evaluation, a net promoter score (NPS (Registered Trademark No. 5028108)).

The rule for recognizing the target group may be based on the answer to the questionnaire different from the questionnaire to be analyzed. Therefore, the rule may be, for example, a rule based on the answer to a specific question as described above, or a rule based on the answer to a plurality of questions. Further, as the question to be adopted at that time, if a question not included in the questionnaire to be analyzed is adopted, a group having less influence from the questionnaire to be analyzed can be targeted, which is more preferable.

The tendency value recognition unit 1j (group tendency value recognition unit, population tendency value recognition unit) refers to the entire population and the target group recognized by the response recognition unit 1h based on the answers recognized by the response recognition unit 1h. , Recognize the population tendency value and group tendency value in the response to the questionnaire different from the questionnaire used when recognizing the target group.

Here, the "trend value" is a value based on the tendency of the response to the questionnaire. For example, the tendency value is a value indicating the ratio of the answer contents, the ratio of the number of people who answered "yes" to the predetermined two-choice question answered with "yes" and "no", and the questionnaire from the start of the questionnaire. Time to completion, percentage of people who gave valid answers, frequency or time of use of customers of the application used when conducting the survey, rate of change in the answer content when the same question is asked at different times, The ratio of each stage in the stage evaluation, the numerical value or the tendency of words included in the answer, etc. can be mentioned.

Further, here, the "population" to be recognized as the population tendency value may be a group that can grasp the rough tendency of the entire population. Therefore, this population includes not only the entire population including the target group, but also all the respondents other than the target group included in the population.

The analysis system S recognizes a group tendency value recognition unit that recognizes a group tendency value that is a value based on the response tendency of the target group, and a population tendency value that is a value based on the response tendency of the entire population. The population inclination recognition unit is configured as one processing unit. However, the present invention is not limited to such a configuration, and the group tendency value recognition unit and the overall inclination value recognition unit may be configured as processing units that are independent of each other.

The divergence degree recognition unit 1k recognizes the divergence degree based on the population tendency value and the group tendency value recognized by the tendency value recognition unit 1j.

Here, the "dissociation degree" is a value indicating the degree of dissociation between the population tendency value and the group tendency value. For example, if the tendency value is a value indicating the ratio of the number of people who answered "yes" to a predetermined two-choice question that answers "yes" and "no", the population tendency value and the group tendency value. The absolute value of the difference from and may be used as the degree of deviation. Further, for example, the absolute value of the difference may be compared with the average value in the population, and the deviation may be used as the degree of deviation. Further, for example, a value obtained by dividing the group tendency value by the population tendency value may be used as the degree of divergence.

Further, the criteria for determining whether or not there is a divergence may be appropriately set by the system designer, and may be determined based on a threshold value artificially set by the system designer himself, for example. The judgment may be made based on a threshold value set based on the result of machine learning using AI or the like.

The singular keyword / tag recognition unit 1l is used to recognize the degree of deviation (that is, the population tendency value and the group tendency value) when the degree of deviation recognized by the degree of deviation recognition unit 1k is equal to or higher than a predetermined threshold value. At least one of the keywords and tags related to the questionnaire is recognized as a peculiar keyword or a peculiar tag for the target group recognized by the target group recognition unit 1i.

After that, the peculiar keyword / tag recognition unit 1l associates the peculiar keyword or the peculiar tag with the target group (that is, the customers constituting the target group) and stores the peculiar keyword or the peculiar tag in the customer data storage unit 1d.

Further, the peculiar keyword / tag recognition unit 1l selects at least one of the keywords and tags related to the questionnaire used for recognizing the degree of deviation when the degree of deviation recognized by the degree of deviation recognition unit 1k is equal to or higher than a predetermined threshold value. , Recognize as a peculiar keyword or peculiar tag for at least one of the tag and the keyword related to the questionnaire used when recognizing the target group.

After that, the peculiar keyword / tag recognition unit 1l stores the peculiar keyword or the peculiar tag in the keyword / tag storage unit 1b in association with at least one of the tag and the keyword related to the questionnaire used when recognizing the target group. ..

Here, the "unique keyword" and the "unique tag" are keywords having specificity to at least one of the target group or the keyword and tag related to the questionnaire used when determining the target group. And tags.

In addition, "specificity" here means a unique tendency and value compared to the reaction data of the entire population when marketing is performed using the target group or the group related to the keyword and tag. It means that there is a high possibility that reaction data with etc. can be obtained.

The match degree recognition unit 1m is associated with at least one of the keywords recognized and the tags given to the questionnaire to be conducted, the group stored in the keyword tag storage unit 1b, or the keywords or tags. Recognize the degree of matching with singular keywords and singular tags.

Here, the "matching degree" may be appropriately set by the system designer, may be artificially set by the system designer himself, or is based on the result of machine learning using AI or the like. It may be set.

For example, as for the degree of matching, simply comparing the keywords and tags related to the questionnaire to be conducted with the keywords and tags related to the group to be compared, or the keywords and tags related to the keywords or tags to be compared, and simply how many matches. You may use a value such as.

Further, for example, the ratio of the number of keywords to the number of tags may be used instead of the number of keywords and tags. Also, for example, the degree of agreement may be recognized based on the deviation of those values from the overall average, rather than directly comparing the values for the group and keywords or tags to be compared with the questionnaire to be conducted.

In the analysis system S, a peculiar keyword and a peculiar tag are stored in association with each customer group in the customer data storage unit 1d. Therefore, when recognizing the degree of matching in the group, the keywords and tags related to the questionnaire to be conducted are compared with the keywords and tags associated with the group, and based on the number of matching keywords and tags. And recognize the degree of agreement.

Then, when the degree of agreement in the group is equal to or greater than a predetermined threshold value, the degree of agreement recognition unit 1m transmits the group from the customer data storage unit 1d to the proposal unit 1e as a group recommending the implementation of the questionnaire. To do. On the other hand, when the degree of agreement in the group is less than a predetermined threshold value, the degree of agreement recognition unit 1m transmits to the proposal unit 1e as a group that does not recommend (exclude) the implementation of the group questionnaire.

Further, in the analysis system S, a peculiar keyword and a peculiar tag are stored in association with each keyword or tag in the keyword / tag storage unit 1b. Therefore, when recognizing the degree of matching in a keyword or tag, the keyword or tag related to the questionnaire to be conducted is compared with the peculiar keyword or peculiar tag associated with the keyword or tag, and the matching keyword or tag is compared. Recognize the degree of agreement based on the number of.

Then, when the degree of matching in the keyword or tag is equal to or higher than a predetermined threshold value, the matching degree recognition unit 1m considers the keyword or tag as a keyword or a keyword or a tag to be considered when determining the population to be surveyed. It is transmitted to the proposal unit 1e as a tag. On the other hand, when the degree of matching in the keyword or tag is less than a predetermined threshold value, the matching degree recognition unit 1m does not need to consider the keyword or tag when determining the population to be surveyed. It is transmitted to the proposal unit 1e as a keyword or tag.

Further, the present invention is not limited to such a configuration, and criteria for recommending and deprecating group proposals based on the degree of agreement may be set as appropriate. For example, if it is proposed to conduct a questionnaire when the degree of agreement is less than the threshold value, new knowledge may be obtained even if reaction data useful for direct marketing cannot be obtained.

The present invention is not limited to such a configuration, and the matching degree recognition unit may be omitted when the proposal regarding the population based on the matching degree is not made to the client.

[Processes executed in each processing unit]
Next, with reference to FIGS. 2 to 5, a process (questionnaire result analysis method) performed when the analysis system S conducts a questionnaire to the customer in response to a request from the client will be described.

In the following explanation, the target group determined based on the answers to the first questionnaire by conducting the first questionnaire on the liquid toothpaste a and the second questionnaire on the camera b, and the first We are analyzing whether or not the keywords and tags related to the second questionnaire have specificity with respect to the keywords and tags related to the questionnaire.

However, the present invention is not only capable of performing such an analysis for each individual questionnaire. Specifically, a large number of questionnaires are conducted, the answers to those questionnaires are accumulated, and at least one of the questionnaires used to determine the target group and the questionnaire regarding keywords or tags for analyzing specificity are repeatedly changed. It may be configured to accumulate a large number of analysis data.

Furthermore, even if the questionnaire to be analyzed and the questionnaire used to recognize the target group are not changed, the questionnaire to be analyzed is performed multiple times while changing the question to be adopted from the questionnaire used to recognize the target group. You may try to analyze.

[Processing when conducting the first questionnaire]
First, with reference to FIGS. 2 and 3, the processing executed by the analysis system S when receiving a request for the first questionnaire (first questionnaire) from the client and executing the first questionnaire will be described.

Here, the first questionnaire is conducted for the purpose of the client manufacturer A wanting to know the image that the customer has for the liquid toothpaste a that he / she manufactures and sells. And.

In this process, first, the questionnaire determination unit 1a determines the questionnaire based on the target information of the client's questionnaire input to the client terminal 2 (FIG. 3 / STEP100).

Specifically, the questionnaire determination unit 1a recognizes the purpose of the client wanting to know the image that the customer has for the liquid toothpaste a based on the information input to the client terminal 2. Then, the questionnaire determination unit 1a refers to the past questionnaires regarding similar products in the past (for example, other liquid toothpastes, oral care products, etc.), recognizes a plurality of questions to be asked to the customer, and implements this time. Consists of 1 questionnaire.

Next, the keyword / tag processing unit 1c executes a process of recognizing a keyword or a process of assigning a tag to the determined first questionnaire (FIG. 3 / STEP101).

Specifically, the keyword / tag processing unit 1c executes text mining on the text of the question constituting the first questionnaire, and recognizes the noun that is the subject of the text as a keyword based on the result. ..

In the present embodiment, the first questionnaire is a questionnaire consisting of a plurality of questions regarding the liquid toothpaste a. Therefore, for the first questionnaire, the keywords are "liquid toothpaste" indicating the category, the product name itself (brand) "a", and the name "A" of the manufacturer that manufactures the liquid toothpaste a. Is recognized as.

Further, the keyword / tag processing unit 1c refers to the data stored in the keyword / tag storage unit 1b and asks a question such as a word indicating a higher-level concept of the recognized keyword, a word generally recalled, or the like. It is given as a tag for each or as a tag for the entire questionnaire.

In the present embodiment, for the entire first questionnaire or for each question constituting the first questionnaire, "oral care products" which is a superordinate concept directly associated with the keyword "liquid toothpaste". Words such as "a", "convenience store" which is the sales channel of the product "a", words such as "drug store", the specific price range of the product "a" (for example, "low price"), and the product "a" The type of promotion route being implemented (for example, "TVCM", "Web advertisement") and the like are added as tags.

Next, the proposal unit 1e proposes to the client a population that can be selected by the client via the client terminal 2 (FIG. 3 / STEP102).

Specifically, first, the proposal unit 1e determines the attributes of the customer based on the type of service received by the client and the like. Next, the proposal unit 1e searches the customer data storage unit 1d for customers corresponding to the attribute, and summarizes the search results as a group for each attribute. After that, the proposal unit 1e displays the group together with the attributes of the group and the number of people included in the group in a format selectable on the client terminal 2.

Next, the population determination unit 1f determines the population based on the group selected by the client via the client terminal 2 (FIG. 3 / STEP103).

Specifically, the population determination unit 1f adds up the groups selected by the client and determines as a population.

Next, the questionnaire implementation unit 1g conducts the determined first questionnaire to the determined population (Fig. 3 / STEP104).

Specifically, the questionnaire implementation unit 1g displays the determined first questionnaire in a format that can be answered to the customer terminal 3 corresponding to the customers included in the determined population.

Next, the answer recognition unit 1h aggregates the answers input via the customer terminal 3 according to a predetermined rule, and transmits the result of the questionnaire to the client terminal 2 (FIG. 3 / STEP105).

Here, the predetermined rule for aggregating the answers may be appropriately set by the system designer, may be artificially set by the system designer himself, or may be set artificially by the system designer, or machine learning using AI. It may be set based on the result or the like.

For example, the predetermined rule for totaling the answers may be a rule based on so-called simple totaling. In that case, the data presented to the client is, for example, the number or ratio of responses to the questionnaire as numerical data or graph data.

Further, for example, the predetermined rule for totaling the answers may be a rule based on so-called cross tabulation. In that case, the data presented to the client is, for example, an item obtained by multiplying the answers to the conducted questionnaire and the answers to other questionnaires as numerical data or graph data.

Next, the response recognition unit 1h stores the first questionnaire conducted and the response to the first questionnaire in the customer data storage unit 1d in association with the customer who conducted the first questionnaire (FIG. 3 / STEP106). This process ends.

[Processing after conducting the questionnaire]
Next, with reference to FIGS. 2 and 4, the process executed by the analysis system S when the customer is surveyed will be described.

In the following description, by executing the processes of STEP100 to STEP106 described above, after conducting the first questionnaire regarding the liquid toothpaste a, similarly, by executing the processes of STEP100 to STEP106 described above, the above-mentioned processes of STEP100 to STEP106 are executed. It is assumed that the second questionnaire regarding the camera b is performed on the population that has conducted the first questionnaire. Then, the following description is a description regarding the process executed after the second questionnaire is carried out.

In addition, as mentioned above, the first questionnaire is conducted for the purpose of the client manufacturer A wanting to know the image that the customer has for the liquid toothpaste a that he / she manufactures and sells. It is assumed that it is a thing.

Therefore, for the first questionnaire, the keywords are "liquid toothpaste" indicating the category, the product name itself (brand) "a", and the name "A" of the manufacturer that manufactures the liquid toothpaste a. Is recognized as.

In addition, for the entire first questionnaire or for each question that composes the first questionnaire, a word such as "oral care products", which is a superordinate concept directly associated with the keyword "liquid toothpaste", It is implemented for words such as "convenience store" and "drug store", which are the sales channels of the product "a", the specific price range of the product "a" (for example, "low price"), and the product "a". It is assumed that a word such as the type of promotion route (for example, "TVCM", "Web advertisement") is attached as a tag.

Further, it is assumed that the second questionnaire is conducted for the purpose of the client manufacturer B wanting to know the image that the customer has for the camera b that he / she manufactures and sells. ..

In addition, for the second questionnaire, the product name itself (brand) indicating the category "camera" and "b", and the name "B" of the manufacturer of the camera b are recognized as keywords. Suppose you are.

In addition, for the entire second questionnaire or for each question that composes the second questionnaire, the word "small household appliances", which is a higher-level concept directly associated with the keyword "camera", "b" The word "home appliance mass retailer" and "specialty store", which are the sales channels of the product "b", the specific price range of the product "b" (for example, "high price"), and the product "b" are implemented. It is assumed that a word such as the type of promotion route (for example, "published in a specialized magazine") is attached as a tag.

In addition, the category, product name, and manufacturer name that are the objects of the questionnaire are different between the first questionnaire and the second questionnaire. Therefore, as a matter of course, the plurality of questions included in the first questionnaire and the plurality of questions included in the second questionnaire are different in at least one.

On the premise of the above, in this process, the response recognition unit 1h first recognizes the responses of the population to the first questionnaire and the second questionnaire (FIG. 4 / STEP200).

Next, the target group recognition unit 1i recognizes the target group from the population based on the response to the first questionnaire (Fig. 4 / STEP201).

Specifically, the target group recognition unit 1i verifies a group of respondents who have given a predetermined answer to a predetermined question included in the first questionnaire in the population by verifying a peculiar keyword and a peculiar tag. Recognize as a target group for.

In the present embodiment, the question "Have you purchased liquid toothpaste a?" Included in the first questionnaire is set as a predetermined question. In addition, the answer "yes" to the predetermined question is the predetermined answer. Therefore, the target group recognition unit 1i sets the group of customers who have given the predetermined response as the target group.

Next, the tendency value recognition unit 1j recognizes the group tendency value of the target group for the second questionnaire based on the response of the target group to the second questionnaire (FIG. 4 / STEP202).

In this embodiment, a group tendency value for verifying the percentage of customers who answered "yes" to the answer to the question "Have you purchased camera b?" Included in the second questionnaire. It is supposed to be.

Therefore, the tendency value recognition unit 1j recognizes a value indicating the ratio of customers who answered "yes" in the target group to the question as a group tendency value. In the present embodiment, the group tendency value is set to 60% (0.60). That is, it is assumed that 60% of the target group answered that they have purchased the camera b.

Next, the tendency value recognition unit 1j recognizes the population tendency value for the second questionnaire based on the response of the entire population to the second questionnaire (FIG. 4 / STEP203).

In this embodiment, as with the group trend value, the percentage of customers who answered "yes" to the answer to the question "Have you purchased camera b?" Included in the second questionnaire. It is the population tendency value.

Therefore, the tendency value recognition unit 1j recognizes a value indicating the ratio of customers who answered “yes” to the question as the population tendency value. In this embodiment, the population tendency value is 20% (0.20). That is, it is assumed that 20% of the entire population answered that they have purchased camera b.

Next, the divergence degree recognition unit 1k recognizes the divergence degree based on the group tendency value and the population tendency value (FIG. 4 / STEP204).

Specifically, the divergence degree recognition unit 1k compares the group tendency value and the population tendency value, and calculates and recognizes the divergence degree. In the present embodiment, the absolute value of the difference between the group tendency value and the population tendency value is used as the degree of divergence. Therefore, in the present embodiment, the degree of divergence is 0.40.

Next, the singular keyword / tag recognition unit 1l determines whether or not the degree of deviation is equal to or greater than a predetermined threshold value (FIG. 4 / STEP205).

Here, the degree of divergence is a value that can be appropriately set by the system designer, may be artificially set by the system designer himself, or is set based on the result of machine learning using AI or the like. May be good. In this embodiment, 0.35 is set as a threshold value.

When the degree of deviation is less than the threshold value (NO in STEP205), the analysis system S ends the current process.

On the other hand, when the degree of divergence is equal to or higher than the threshold value (YES in STEP205), the singular keyword / tag recognition unit 1l recognizes the keyword and tag related to the second questionnaire as the singular keyword and the singular tag (FIG. 4 / STEP206).

In the present embodiment, the degree of deviation is 0.40, which exceeds a predetermined threshold value (0.35). Therefore, the peculiar keyword / tag recognition unit 1l recognizes all the keywords and tags related to the second questionnaire as peculiar keywords or peculiar tags.

In the present embodiment, keywords such as "camera", product name "b", and manufacturer name "B" are recognized as peculiar keywords. In addition, words such as "small home appliances", "home appliance mass retailers", "specialty stores", "high prices", and "specialized magazine publication" are recognized as peculiar tags.

Next, the peculiar keyword / tag recognition unit 1l associates the peculiar keyword and the peculiar tag with the customer of the target group and stores them in the customer data storage unit 1d (FIG. 4 / STEP207).

In this embodiment, the customer who answered "yes" to the question "Have you purchased liquid toothpaste a?" Included in the first questionnaire with the above-mentioned peculiar keyword and peculiar tag. Associate with and store.

Next, the peculiar keyword / tag recognition unit 1l stores the peculiar keyword and the peculiar tag in the keyword / tag storage unit 1b in association with the keyword and the tag related to the first questionnaire (FIG. 4 / STEP208), and this time processing. To finish.

In the present embodiment, the above-mentioned peculiar keywords and peculiar tags are used as keywords such as "liquid toothpaste", product name "a", manufacturer name "A", and "oral care products", "convenience store", and "drug store". , "Low price", "TVCM", "Web advertisement" and so on.

[Processing when conducting the second and subsequent questionnaires]
Next, with reference to FIGS. 2 and 5, when the analysis system S receives a request for a second questionnaire from a client, a process to be executed before the questionnaire is performed will be described.

In the following description, it is a process when the third questionnaire is carried out. Here, the third questionnaire is conducted for the purpose of the client manufacturer C wanting to know the image that the customer has for the watch c that he / she manufactures and sells. To do.

On the premise of the above, in this process, first, the questionnaire determination unit 1a determines the questionnaire based on the target information of the client's questionnaire input to the client terminal 2 (FIG. 5 / STEP300).

Next, the keyword / tag processing unit 1c executes a process of recognizing a keyword or a process of assigning a tag to the determined first questionnaire (FIG. 5 / STEP301).

In these STEP300 and STEP301, the questionnaire determination unit 1a and the keyword / tag processing unit 1c execute the same processing as those of STEP100 and STEP101 described above.

In the present embodiment, for the third questionnaire, the product name itself (brand) indicating the category "watch", "c", the name "C" of the manufacturer of the watch c, etc. are used. , Suppose it is recognized as a keyword.

In addition, for the entire third questionnaire or for each question that composes the third questionnaire, the word "small household appliances", which is a higher-level concept directly associated with the keyword "clock", "c" The word "home appliance mass retailer" and "specialty store", which are the sales channels of the product "c", the specific price range of the product "c" (for example, "high price"), and the product "c" are implemented. It is assumed that a word such as the type of promotion route (for example, "magazine publication") is attached as a tag.

Next, the matching degree recognition unit 1m recognizes the matching degree between the keyword and the assigned tag recognized for the third questionnaire and the peculiar keyword and the peculiar tag stored in the keyword tag storage unit 1b. (Fig. 5 / STEP302).

Specifically, the matching degree recognition unit 1m includes a peculiar keyword and a peculiar tag associated with each group of customers designated as the target group in the processes of STEP200 to STEP207 described with reference to FIG. , The number of keywords and tags common to them is recognized as the degree of agreement by comparing with the keywords and tags related to the third questionnaire.

In addition, the matching degree recognition unit 1m is a unique keyword associated with the keyword or tag for each of the keywords and tags related to the first questionnaire in the processes of STEP200 to STEP206 and STEP207 described with reference to FIG. The singular tag is compared with the keyword and tag related to the third questionnaire, and the number of keywords and tags common to them is recognized as the degree of matching.

Next, the proposal unit 1e proposes a group recommended to be included in the population and a group recommended to be excluded from the population via the client terminal 2 based on the degree of agreement (FIG. 5). / STEP303).

Specifically, the proposal unit 1e recommends that the population include a group in which the degree of agreement recognized by the degree of agreement recognition unit 1m is equal to or higher than the threshold value (4 in the present embodiment) for recommendation. Is recognized as, and displayed on the client terminal 2 in a selectable format.

In addition, the proposal unit 1e recognizes as a group that recommends excluding the group whose degree of agreement recognized by the degree of agreement recognition unit 1m is less than the threshold value (2 in this embodiment) for not recommending it from the population. Then, it is displayed on the client terminal 2 in a selectable format.

At this time, the proposal unit 1e displays the attributes of the group in a form attached to the group together with the group displayed in the selectable format on the client terminal 2.

In addition to the group recommended to be included in the population and the group recommended to be excluded from the population, the proposal unit 1e can select to determine the population regardless of those groups. indicate.

In the present embodiment, when the keywords and tags related to the third questionnaire are compared with the peculiar keywords and peculiar tags recognized in the processes of STEP200 to STEP207 described with reference to FIG. 4, "small home appliances" and "home appliance mass sales" are compared. Tags such as "store", "specialty store", and "high price" match. Therefore, the group with the peculiar keyword and the peculiar tag is a group recommended to be included in the population.

Next, the population determination unit 1f determines whether or not the population has been determined based on the information input from the client to the client terminal 2 (FIG. 5 / STEP304).

Specifically, the population determination unit 1f determines that when one group recommended to be included in the population is selected, that group is determined as the population, and when a plurality of groups are selected. Judges that the sum of these groups was determined as the population. On the other hand, if it is selected to determine the population regardless of the proposed group, it is determined that the population has not been determined.

When the population is not determined (NO in STEP 304), the proposal unit 1e presents a group that the client can select as a population and keywords and tags related to the group via the client terminal 2. (Fig. 5 / STEP305).

Specifically, first, the proposal unit 1e determines the attributes of the customer based on the type of service received by the client and the like. Next, the proposal unit 1e searches the customer data storage unit 1d for customers corresponding to the attribute, and summarizes the search results as a group for each attribute. After that, the proposal unit 1e displays the group together with the attributes of the group and the number of people included in the group in a format selectable on the client terminal 2.

At this time, the proposal unit 1e displays in a selectable format that the population is determined regardless of the groups in addition to the groups that the client can select as the population.

Further, the proposal unit 1e displays the keywords and tags associated with the customers included in the group on the client terminal 2 together with the group in descending order of the number.

Next, the proposal unit 1e proposes keywords and tags to be considered when determining the population based on the degree of agreement (FIG. 5 / STEP306).

Specifically, the proposal unit 1e should consider keywords and tags whose matching degree recognized by the matching degree recognition unit 1m is equal to or higher than a predetermined threshold value (that is, effective for collecting reaction data). Is recognized as, and is displayed on the client terminal 2.

Further, the proposal unit 1e does not need to particularly consider a group in which the degree of agreement recognized by the degree of agreement recognition unit 1m is less than a predetermined threshold value (that is, it does not significantly affect the collection of reaction data) as a keyword or tag. Recognize and do not display on client terminal 2.

Next, the population determination unit 1f determines whether or not the population has been determined based on the information input from the client to the client terminal 2 (FIG. 5 / STEP307).

Specifically, when one group is selected from the plurality of groups presented in STEP 305, the population determination unit 1f determines that the group has been determined as the population, and multiple selections are made. In some cases, it is judged that the sum of these groups is determined as the population. On the other hand, if it is selected to determine the population regardless of those groups, it is determined that the population has not been determined.

If the population is not determined (NO in STEP308), Proposal 1e changes the threshold for concordance (FIG. 5 / STEP308).

Specifically, the proposal unit 1e lowers the threshold value for determining the group recommended to be included in the population. In this embodiment, it is changed from 4 to 3. It also lowers the threshold for determining which groups are recommended to be excluded from the population. In this embodiment, it is changed from 2 to 1. It also lowers the threshold for determining keywords or tags to consider.

After that, the analysis system S re-executes the processes of STEP 303 to STEP 307. When the same process is executed by changing the threshold value to be low in this way, the proposed group, the group recognized as a keyword or tag, and the keyword or tag increase. Therefore, the population is more likely to be determined.

On the other hand, when the population is determined (Y in STEP304 or STEP307), the questionnaire implementation unit 1g conducts the determined first questionnaire on the determined population (FIG. 5 / STEP309).

Next, the answer recognition unit 1h aggregates the answers input via the customer terminal 3 according to a predetermined rule, and transmits the result of the questionnaire to the client terminal 2 (FIG. 5 / STEP310).

Next, the response recognition unit 1h stores the first questionnaire conducted and the response to the first questionnaire in the customer data storage unit 1d in association with the customer who conducted the first questionnaire (FIG. 5 / STEP311). This process ends.

In these STEP309 to STEP311 the analysis system S executes the same processing as the above-mentioned STEP104 to STEP106.

As described above, the analysis system S indicates the process of recognizing the keywords included in the questions constituting the questionnaire and the attributes of the questions for the first questionnaire and the second questionnaire. The process of adding a tag is being executed. That is, in the analysis system S, keywords or tags (and thus, the field of marketing) that can be examined from the answers of the first questionnaire and the second questionnaire are set.

After that, the analysis system S recognizes the target group from the population based on the response to the first questionnaire, and the tendency value of the target group and the tendency of the entire population based on the response to the second questionnaire. Recognize the degree of deviation from the value.

After that, the analysis system S determines whether or not the keyword or tag related to the second questionnaire has specificity to the target group and the keyword or tag related to the first questionnaire based on the degree of deviation. Recognize and store them in association with each other if they have specificity.

Then, the analysis system S proposes a group recognized based on at least one of the singular keyword and the singular tag as a population for carrying out the third questionnaire. If the population is not determined based on the proposal, at least one of the singular keyword and the singular tag is proposed as a keyword or tag to be considered when determining the population to be subjected to the third questionnaire. To do.

Here, the peculiar keyword or peculiar tag is specific to a certain group or the keyword or tag related to the first questionnaire, and the keyword or tag related to the third questionnaire is a peculiar keyword or peculiarity. If it matches the tag, it is highly probable that the keyword or tag related to the third questionnaire also has specificity for the group or the keyword or tag related to the first questionnaire.

That is, in such a case, the keyword or tag related to the group or the first questionnaire is likely to be a group or keyword or tag having a peculiar tendency also in the field of marketing related to the third questionnaire. Is high.

Therefore, when a third survey is conducted on a group or a population determined using keywords or tags, a specific answer to the third survey (that is, marketing related to the third survey) Reaction data) that is highly useful in the field can be obtained.

Although the illustrated embodiment has been described above, the present invention is not limited to such an embodiment.

For example, in the above embodiment, the same population is used as the population that carries out the first questionnaire and the population that carries out the second questionnaire. However, the questionnaire result analysis system and the questionnaire result analysis method of the present invention are not limited to such a configuration, and it is sufficient that the group to be verified is included in both populations.

Further, in the above embodiment, when recognizing the population tendency value, the response of the entire population to the second questionnaire is used. However, the questionnaire result analysis system and the questionnaire result analysis method of the present invention are not limited to such a configuration, and may be a group capable of grasping a rough tendency of the entire population.

Therefore, for example, the population used to recognize the population tendency value may be all respondents other than the target group included in the population.

Further, in the above embodiment, the peculiar keyword and the peculiar tag are stored in association with all the keywords and tags related to the customer of the target group and the first questionnaire. However, the present invention is not limited to such a configuration.

Therefore, for example, the peculiar keyword and the peculiar tag may be stored in association with the target group instead of the individual customer. Further, the peculiar keyword and the peculiar tag may be stored in association with only the keyword and the tag related to the question used when recognizing the target group.

Further, in the above embodiment, the population is determined when the population is not determined after proposing a group to be recommended to be included in or excluded from the population based on the singular keyword or the singular tag. It proposes keywords and tags that should be considered when doing so.

However, the present invention is not limited to such a configuration. For example, one or two of the groups to be included in the population, the keywords to be considered, and the tags to be considered may be proposed. In addition, the order of proposals may be changed as appropriate, and groups and keywords and tags may be proposed at the same time.

1 ... server, 1a ... questionnaire determination unit, 1b ... keyword / tag storage unit, 1c ... keyword / tag processing unit (first keyword / tag processing unit, second keyword / tag processing unit, third keyword / tag processing unit) 1, 1d ... Customer data storage unit, 1e ... Proposal unit (population proposal unit, keyword / tag proposal unit), 1g ... Questionnaire implementation unit, 1h ... Answer recognition unit (first answer recognition unit, second answer recognition unit), 1i ... Target group recognition unit, 1j ... Tendency value recognition unit (group tendency value recognition unit, population tendency value recognition unit), 1k ... Deviation degree recognition unit, 1l ... Singular keyword / tag recognition unit, 1m ... Matching degree recognition unit 2, 2 ... Client terminal, 3 ... Customer terminal, S ... Analysis system (questionnaire result analysis system).

Claims (6)

It is a questionnaire result analysis system for analyzing a population consisting of a plurality of respondents who answered the above questions based on the answers to the questions constituting the questionnaire.
A second questionnaire consisting of a plurality of questions containing at least one question different from the plurality of questions constituting the first questionnaire is based on the questions included in the second questionnaire. , A process of recognizing a keyword included in the question as a keyword related to the second questionnaire, and a process of assigning a tag indicating the attribute of the question to the second questionnaire as a tag related to the second questionnaire. A second keyword / tag processing unit that executes at least one,
The first answer recognition unit that recognizes the answer by the population to the first questionnaire, and
The second answer recognition unit that recognizes the answer by the population to the second questionnaire, and
Recognizing target group recognition unit as the target group the group that was given the answer against a predetermined questions contained in the first survey of the population,
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, and a value based on the tendency of the response of the target group to the second questionnaire. The group tendency value recognition unit that recognizes as the group tendency value that is
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population, and a value based on the tendency of the population to answer the second questionnaire. The population inclination recognition unit that recognizes as the population tendency value , which is
A divergence degree recognition unit that recognizes the degree of divergence, which is a value indicating the degree of divergence between the group tendency value and the population tendency value based on the group tendency value and the population tendency value.
When the degree of divergence is equal to or greater than a predetermined threshold value, at least one of the keyword and the tag related to the second questionnaire is a specific keyword or a tag that is specific to the target group. Alternatively, a questionnaire result analysis system characterized by having a peculiar keyword / tag recognition unit that recognizes as a peculiar tag. In the questionnaire result analysis system according to claim 1,
A third questionnaire composed of a plurality of questions constituting the first questionnaire and a plurality of questions including at least one question different from the plurality of questions constituting the second questionnaire. On the other hand, based on the question included in the third questionnaire, the process of recognizing the keyword contained in the question as a keyword related to the third questionnaire, and the tag indicating the attribute of the question are attached. A third keyword / tag processing unit that executes at least one of the processes assigned to the third questionnaire as a tag related to the third questionnaire, and
A matching degree recognition unit that recognizes the degree of matching between at least one of the peculiar keyword and the peculiar tag and at least one of the keyword and the tag related to the third questionnaire.
When the degree of agreement is equal to or higher than a predetermined threshold value, the target group is proposed as a population to carry out the third questionnaire, or is proposed as a group to be excluded from the population to be carried out. Questionnaire result analysis system characterized by having. It is a questionnaire result analysis method for analyzing a population consisting of a plurality of respondents who answered the above based on the answers to the questions constituting the questionnaire.
In contrast to the second questionnaire, in which the second keyword / tag processing unit is composed of a plurality of questions including at least one question different from the plurality of questions constituting the first questionnaire, the second questionnaire is described. Based on the question contained in, the process of recognizing the keyword contained in the question as a keyword related to the second questionnaire, and the tag indicating the attribute of the question as the tag related to the second questionnaire. The process of executing at least one of the processes given to the second questionnaire, and
The process in which the first answer recognition unit recognizes the answer by the population to the first questionnaire,
The process in which the second answer recognition unit recognizes the answer by the population to the second questionnaire,
Target group recognition unit, and a step of recognizing a target group of the group performing a predetermined answer by pairs in a predetermined questions contained in the first survey of the population,
The group tendency value recognition unit sets a value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, to the target group for the second questionnaire. The process of recognizing as a group tendency value , which is a value based on the tendency of the answer of
The value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population by the population inclination recognition unit with respect to the second questionnaire. The process of recognizing as a population tendency value , which is a value based on the population's response tendency,
A step of recognizing the degree of dissociation, which is a value indicating the degree of dissociation between the group tendency value and the population tendency value, based on the group tendency value and the population tendency value.
When the degree of deviation is equal to or greater than a predetermined threshold value, the singular keyword / tag recognition unit has specificity for the target group for at least one of the keyword and the tag related to the second questionnaire. A method for analyzing a questionnaire result, which comprises a step of recognizing a peculiar keyword or a peculiar tag which is a keyword or a tag. A questionnaire result analysis system for analyzing the questionnaire based on the answers to the questions that make up the questionnaire.
For a first questionnaire composed of a plurality of questions, a process of recognizing a keyword included in the question as a keyword related to the first questionnaire based on the question included in the first questionnaire. , And a first keyword / tag processing unit that executes at least one of the processes of assigning a tag indicating the attribute of the question to the first questionnaire as a tag related to the first questionnaire.
Based on the questions included in the second questionnaire, as opposed to the second questionnaire composed of a plurality of questions containing at least one question different from the plurality of questions constituting the first questionnaire. The process of recognizing the keyword included in the question as a keyword related to the second questionnaire, and the process of assigning a tag indicating the attribute of the question to the second questionnaire as a tag related to the second questionnaire. A second keyword / tag processing unit that executes at least one of
The first answer recognition unit that recognizes the answers by the population to the first questionnaire, and
The second answer recognition unit that recognizes the answer by the population to the second questionnaire, and
Recognizing target group recognition unit as the target group the group that was given the answer against a predetermined questions contained in the first survey of the population,
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, and a value based on the tendency of the response of the target group to the second questionnaire. The group tendency value recognition unit that recognizes as the group tendency value that is
A value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population, and a value based on the tendency of the population to answer the second questionnaire. The population inclination recognition unit that recognizes as the population tendency value , which is
A divergence degree recognition unit that recognizes the degree of divergence, which is a value indicating the degree of divergence between the group tendency value and the population tendency value based on the group tendency value and the population tendency value.
When the degree of divergence is equal to or greater than a predetermined threshold value, the specificity of at least one of the keyword and the tag related to the second questionnaire is given to at least one of the keyword and the tag related to the first questionnaire. A questionnaire result analysis system including a peculiar keyword / tag recognition unit that recognizes a peculiar keyword or a peculiar tag that is a possessed keyword or tag. In the questionnaire result analysis system according to claim 4,
A third questionnaire composed of a plurality of questions constituting the first questionnaire and a plurality of questions including at least one question different from the plurality of questions constituting the second questionnaire. On the other hand, based on the question constituting the third questionnaire, the process of recognizing the keyword included in the question as the keyword related to the third questionnaire, and the tag indicating the attribute of the question are attached. A third keyword / tag processing unit that executes at least one of the processes assigned to the third questionnaire as a tag related to the third questionnaire, and
A matching degree recognition unit that recognizes the degree of matching between at least one of the peculiar keyword and the peculiar tag and at least one of the keyword and the tag related to the third questionnaire.
When the degree of agreement is equal to or higher than a predetermined threshold value, at least one of the peculiar keyword and the peculiar tag is proposed as a keyword or tag to be considered when determining a population for conducting the third questionnaire. Questionnaire result analysis system characterized by having a keyword / tag proposal department. A questionnaire result analysis method for analyzing the questionnaire based on the answers to the questions that make up the questionnaire.
For the first questionnaire composed of a plurality of questions, the first keyword / tag processing unit sets the keywords included in the question based on the questions included in the first questionnaire. A step of executing at least one of a process of recognizing as a keyword related to the questionnaire 1 and a process of assigning a tag indicating the attribute of the question to the first questionnaire as a tag related to the first questionnaire.
For the second questionnaire, the second keyword / tag processing unit is composed of a plurality of questions including at least one question different from the plurality of questions constituting the first questionnaire. Based on the question included in the questionnaire, the process of recognizing the keyword contained in the question as the keyword related to the second questionnaire, and the tag indicating the attribute of the question as the tag related to the second questionnaire. A step of executing at least one of the processes given to the second questionnaire, and
The process in which the first answer recognition unit recognizes the answers by the population to the first questionnaire,
The process in which the second answer recognition unit recognizes the answer by the population to the second questionnaire,
Target group recognition unit, and a step of recognizing a target group of the group performing a predetermined answer by pairs in a predetermined questions contained in the first survey of the population,
The group tendency value recognition unit sets a value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire among the target groups, to the target group for the second questionnaire. The process of recognizing as a group tendency value , which is a value based on the tendency of the answer of
The value based on the ratio of the number of people who answered a predetermined question to a predetermined question included in the second questionnaire in the population by the population inclination recognition unit with respect to the second questionnaire. The process of recognizing as a population tendency value , which is a value based on the population's response tendency,
A step of recognizing the degree of dissociation, which is a value indicating the degree of dissociation between the group tendency value and the population tendency value, based on the group tendency value and the population tendency value.
When the degree of divergence is equal to or greater than a predetermined threshold value, the singular keyword / tag recognition unit can use at least one of the keyword and the tag related to the second questionnaire to the keyword and the tag related to the first questionnaire. A method for analyzing a questionnaire result, which comprises a step of recognizing as a peculiar keyword or a peculiar tag which is a keyword or tag having peculiarity to at least one of them.

Images