JP4451354B2 - Topic scale management device - Google Patents

Description

  The present invention relates to a document processing technique, and more particularly to a technique for evaluating document contents based on a predetermined evaluation criterion.

In recent years, with the spread of computers and the advancement of network technology, the exchange of electronic information via the network has become popular. An example of this is the weblog, also known as a blog. A user can easily transmit his / her opinion by opening a personal page on a site that provides a weblog service. Due to the ease of opening procedures, the number of web log openers has been increasing steadily, and weblogs are attracting attention as a new information transmission means.
Japanese Patent Laid-Open No. 2003-196301 Manabu Okumura, Yasuyuki Minamino, Yasuaki Fujiki, Yasuhiro Suzuki, “Text Mining Based on Automatic Collection and Monitoring of Blog Pages”, Japan, Japan Society for Artificial Intelligence, July 2, 2004, SIG-SWO-A401-01

  By the way, as a traditional method of collecting information in public opinion surveys and marketing, a method called a questionnaire is generally used. In order for the questionnaire system to function as an effective information collection system, it is natural that the user's willingness to answer or the user must be forced to answer. If the user's willingness to answer the questionnaire is not aroused, the questionnaire system will not function effectively. In the modern society where individuals are connected via the Internet, companies are trying to collect opinions from users by opening a homepage for soliciting opinions and use it for marketing. However, this is also a questionnaire via a network, and it contains the same issues as a paper-based questionnaire.

  The present invention was completed based on the inventor's attention to the characteristics and future potential of "private" information transmission means such as weblogs, and the main purpose of the present invention is to track topic trends via the network. It is to provide a technique for efficient analysis.

One aspect of the present invention is a topic scale management apparatus.
This apparatus collects electronic document data such as web logs. A set of document data to be surveyed is specified according to a survey condition designated by the user, and the topic scale and trend in the set are quantified as a topic scale value and an overall evaluation value, respectively. Then, by generating a figure such as a graph using these values as parameters, the scale and tendency of the topic in the document data set satisfying the desired investigation condition are visually displayed.

  It should be noted that any combination of the above-described components and a representation of the present invention by a method, apparatus, system, recording medium, and computer program are also effective as an aspect of the present invention.

  According to the present invention, a technique for efficiently analyzing topic trends can be provided.

  In the present embodiment, a topic scale management apparatus 100 that collects various articles from a weblog site and analyzes the trend of public topics will be described.

FIG. 1 is a hardware configuration diagram of a topic scale management system.
The topic scale management apparatus 100 collects article data from a plurality of sites (hereinafter simply referred to as “web log site 300”) such as the web log site 300a and the web log site 300b connected via the Internet 12. The weblog site may be existing. The person who establishes the web log can write various opinions as articles on his / her web log site and transmit it on the Internet 12 as usual without being aware of the existence of the topic scale management apparatus 100. From a computer connected to the Internet 12, it is possible to browse the opinions of the web log site thus established, or to write comments on articles.
The topic scale management apparatus 100 and the web log site 300 connected by the Internet 12 constitute the topic scale management system 10 in this embodiment.

The main purpose of the topic scale management apparatus 100 is to grasp the trend in the world from information transmitted by individuals on the network such as web logs. The topic scale management apparatus 100 collects article data from the weblog site 300 and evaluates it from two viewpoints.
(1) Objective viewpoint of what the topic is (hereinafter also referred to as “first viewpoint”)
(2) Subjective viewpoint of what opinions the contributor expresses regarding the topic (hereinafter also referred to as “second viewpoint”)
These are two.
The topic scale management apparatus 100 sets a category in which topic contents are categorized in advance, such as “music” and “entertainment”. The topic scale management device 100 classifies a large amount of article data collected from the weblog site 300 into one of the categories by natural language processing. The topic scale management apparatus 100 also has a function of grouping article data from sentence components, such as a collection of article data including a certain term. In any case, the topic scale management device 100 classifies article data according to the content. This is an outline of processing based on the first viewpoint.
The topic scale management apparatus 100 determines whether or not the content of each article data meets a predetermined evaluation standard. For example, evaluation criteria such as “whether it is a positive content”, “a negative content”, “an opinion based on experience”, or “an opinion based on guess” are assumed. The topic scale management apparatus 100 determines by natural language processing whether or not a large amount of article data collected from the web log site 300 meets the evaluation criteria. This is processing based on the second viewpoint.
Lastly, the topic scale management apparatus 100 quantifies the topic scale and topic trend indicating what kind of content is a large amount of article data and what opinions are expressed, and visually displays the result. In this way, the user of the topic scale management apparatus 100 visually recognizes the topic scale and trend in the set of article data to be investigated. Hereinafter, when summarizing the topic scale and trend, it is also simply referred to as “topic trend”.
Hereinafter, the topic scale management apparatus 100 according to the present embodiment will be described focusing on details of the topic trend analysis and display method.

FIG. 2 is a functional block diagram of the topic scale management apparatus.
Each block shown here can be realized in hardware by an element such as a CPU of a computer or a mechanical device, and in software it is realized by a computer program or the like. Draw functional blocks. Therefore, those skilled in the art will understand that these functional blocks can be realized in various forms by a combination of hardware and software.
Here, functions to be exhibited by each functional block will be mainly described, and specific actions will be described with reference to FIG.

The topic scale management apparatus 100 includes a user interface processing unit 110, a communication processing unit 120, a data processing unit 130, and a data storage unit 160.
The user interface processing unit 110 is in charge of processing related to the entire user interface such as input processing from the user and information display for the user. The communication processing unit 120 is in charge of communication processing with other computers such as the web log site 300 via the Internet 12. The communication processing unit 120 collects article data electronically posted on the weblog site 300.
In this embodiment, the user interface processing unit 110 will be described as being provided with the user interface service of the topic scale management apparatus 100. As another example, the user may operate the topic scale management apparatus 100 via the Internet 12. In this case, the communication processing unit 120 receives operation instruction information from the user terminal and transmits processing result information executed based on the operation instruction to the user terminal.

The data processing unit 130 executes various types of data processing based on input operations via the user interface processing unit 110 and article data acquired from the communication processing unit 120. The data processing unit 130 also serves as an interface among the user interface processing unit 110, the communication processing unit 120, and the data storage unit 160.
The data storage unit 160 stores various setting data prepared in advance and various data including article data received from the data processing unit 130.

The user interface processing unit 110 includes an input unit 112 and a display unit 114. The input unit 112 receives an input operation from the user. The display unit 114 displays various information to the user.
The communication processing unit 120 includes an article data receiving unit 122. The article data receiving unit 122 acquires article data from the weblog site 300. The data storage unit 160 holds in advance a URL (Uniform Resource Locator) list of web log sites to be collected. The article data receiving unit 122 may acquire the URL of the web log from the electronic bulletin board or the link page of the web log, and update the URL list. The article data receiving unit 122 detects the update of the article content by the web log update notification service, and appropriately acquires the article data to be updated. Alternatively, the article data receiving unit 122 may periodically visit the web log site 300 listed in the URL list and collect new article data among the electronically published article data. The article data is subjected to natural language processing by the data processing unit 130 and is classified into one of a plurality of predetermined categories. Categories will be described in more detail in connection with FIG.

The data storage unit 160 includes an article data holding unit 162 and a topic information holding unit 164.
The article data holding unit 162 stores the article data received from the web log site 300 by the article data receiving unit 122. The article data holding unit 162 in this embodiment holds all the collected article data. As a modification, the article data holding unit 162 may hold partial information of article data such as an article title, and may be held together with the URL of the web log site 300 from which the article data is collected. In this case, once the data processing unit 130 causes the communication processing unit 120 to retrieve the original article data from the URL, the article data once collected can be reacquired.
The article data holding unit 162 stores topic scale information indicating the topic scale for each category and overall evaluation information indicating the overall topic trend. The data structure of the topic information holding unit 164 will be described in detail with reference to FIG. How much the topic of the category is popular in the world, that is, the size of the topic scale is quantified as a topic scale value. In addition, what tendency can be read for the entire topic of the category, that is, the overall topic trend viewed from the content of the document data is quantified as an overall evaluation value.

The data processing unit 130 includes a statistical unit 140, an evaluation criterion setting unit 146, and an evaluation unit 150.
The evaluation criterion setting unit 146 sets an evaluation condition that is a criterion for evaluating the content of the article data. Corresponding to the second viewpoint described above, it can be said that this is an evaluation standard for judging the opinion of the poster contributor. The evaluation unit 150 analyzes the content of the article data received by the article data reception unit 122, classifies the article data into any of predetermined categories, and evaluates the article data according to the evaluation condition. The statistics unit 140 calculates a topic scale value and an overall evaluation value for a predetermined article data set.

The evaluation unit 150 includes a range identification unit 152, a survey determination unit 153, a survey condition setting unit 154, an information amount calculation unit 155, a conformity evaluation unit 156, a term detection unit 158, and a classification processing unit 159.
The range specifying unit 152 specifies a data range to be evaluated among the article data acquired from the weblog site 300 by the article data receiving unit 122. The article data may include data unnecessary for topic scale analysis, such as images and advertisements. The range specifying unit 152 specifies a range to be evaluated by excluding such data. A specific processing method will be described with reference to FIG. Hereinafter, the range of data to be evaluated among the article data is referred to as “unit data”.

  The information amount calculation unit 155 calculates the information amount of unit data. For example, the number of characters included in the unit data may be specified as the information amount of the article as it is. In addition, the information amount of the unit data may be calculated, for example, as a logarithmic value with 10 as the base and the number of characters as a true number. In this embodiment, the information amount of unit data is calculated as the square root of the number of characters. For example, in the case of 121 character unit data, the amount of information is 11. In any case, the information amount of the unit data is preferably calculated by a predetermined monotonically increasing function using the number of characters of the unit data as a variable.

  The conformity evaluation unit 156 refers to the evaluation condition by the evaluation criterion setting unit 146 and indexes the degree of conformity of the contents of the unit data as the conformity evaluation value. In this embodiment, the conformity evaluation unit 156 calculates the conformity evaluation value by applying the feature vector method. In the feature vector method, the importance of various words is specified for each evaluation condition. High importance is set for a word representing a characteristic according to the evaluation condition. For example, an affirmative article may contain many terms such as “good”, “good”, “delicious”, “inspiring”, “excellent”, and “effective”. Therefore, these terms are terms having high importance in determining a positive article. On the other hand, in the case of a negative article, it is considered that many terms such as “bad”, “bad”, “bad”, “miss”, and “insufficient” are included. Therefore, these terms are terms having high importance in determining a negative article. The conformity evaluation unit 156 evaluates whether each term included in the unit data is an affirmative term or a negative term, and determines whether the unit data as a whole is positive or negative. It is indexed by the value. The size of this feature vector indicates the degree of fitness of unit data for the evaluation condition. The conformity evaluation unit 156 in the present embodiment normalizes the size of this feature vector to a range of 0 to 100 and calculates a conformity evaluation value. It can be said that the greater the conformity evaluation value, the greater the degree of conformance.

The conformity evaluation unit 156 in the present embodiment can also evaluate unit data from various angles based on evaluation conditions specific to the category. For example, in the case of automobile evaluation, topic trends can be determined for evaluation items unique to automobiles in addition to the general evaluation conditions of positive and negative. Specifically, evaluation items such as “price”, “performance”, and “design” are set. At this time, the conformity evaluation unit 156 first specifies not only what the topic is, but also what the evaluation item is as a first viewpoint. The topic specification for the evaluation item may also be made based on the feature vector method. In this case, a word with high importance and a word that is not so may be set for each evaluation item. For example, a topic related to “price” may include many terms such as “yen”, “high”, and “cheap”. Terms such as “safety”, “displacement”, and “starting on a slope” may be important terms for identifying the topic of “performance”. The conformity evaluation unit 156 specifies what evaluation item the document content that is the subject of the investigation is, by calculating a feature vector for each evaluation item.
With the above processing method, the conformity evaluation unit 156 can calculate a plurality of types of conformity evaluation values corresponding to a plurality of evaluation items even with one unit data.

  The classification processing unit 159 classifies the unit data into any category. The classification processing unit 159 also applies the feature vector method to classify each unit data into any category. The importance of various words is set for each category. High importance is set for words representing the characteristics of the category. Depending on how many words that are important in each category are included in the unit data to be classified, how much content is appropriate for each category is indexed by a feature vector. The size of the feature vector represents the degree of matching between the category and the unit data to be classified. The classification processing unit 159 in this embodiment normalizes the size of the feature vector to a range of 0 to 100 and calculates it as a category evaluation value. It can be said that the greater the category evaluation value, the greater the fitness. The classification processing unit 159 calculates a category fitness value for each unit data, and classifies the unit data into a category having the highest fitness.

  The survey condition setting unit 154 sets survey conditions for specifying article data to be surveyed when surveying topic trends. The user can set the condition to be satisfied by the article data to be investigated as the investigation condition. For example, it is possible to investigate the scale and tendency of the topic with respect to topics related to three vehicle types, “vehicle type A, vehicle type B, and vehicle type C”. The investigation determination unit 153 determines whether or not each received unit data should be an object of investigation according to the set investigation condition. The term detection unit 158 detects subordinate terms and equivalent terms of terms specified by the user. The purpose and principle will be described in detail with reference to FIG.

The statistic unit 140 includes a topic scale value calculation unit 142 and an overall evaluation value calculation unit 144.
The topic scale value calculation unit 142 calculates a topic scale value in a predetermined set of article data. For example, when unit data of information amount A and information amount B is classified into category 1, the topic scale value related to category 1 is obtained as information amount A + information amount B. In this way, the topic scale value calculation unit 142 calculates the topic scale value for each category. The more the article data classified into category 1, the larger the topic scale value. Also, the topic scale value increases as the amount of information included in the article data increases. Rather than simply calculating the topic scale value based on the number of articles and the number of words, the topic scale value is calculated using the number of articles and the information amount of each article as variables. Further, the topic scale value calculation unit 142 compares the topic scale values in one set of article data and another set of article data, and calculates a relative value thereof. Hereinafter, when the topic scale value as an absolute value is distinguished from the topic scale value as a relative value, the former is referred to as an “absolute topic scale value” and the latter is referred to as a “relative topic scale value”. The topic scale value calculation unit 142 calculates the topic scale value not only for a set of article data corresponding to a category but also for a set of article data including a specific term, which will be described in detail later.

  The overall evaluation value calculation unit 144 calculates an overall evaluation value for a predetermined set of article data. For example, when the unit data of the conformity evaluation value 70 and the conformity evaluation value 60 are classified into category 1 as the degree of a positive article, the overall evaluation value for category 1 is obtained as (70 + 60) / 2 = 65. . In this way, the overall evaluation value calculation unit 144 calculates the overall evaluation value from the conformity evaluation value. The overall evaluation value calculation unit 144 may calculate the overall evaluation value by performing a weighted average of the conformity evaluation values of the unit data based on the information amount of the unit data. The conformity evaluation unit 156 may calculate the conformity evaluation value of the unit data as a binary value depending on whether the article is a positive article or a negative article. In this case, the overall evaluation value calculation unit 144 may use a ratio between the number of positive article data and the number of negative article data for a predetermined set of article data as the overall evaluation value. Alternatively, a ratio between the topic scale value as the entire positive article data and the topic scale value as the entire negative article data may be calculated as the overall evaluation value. In any case, the overall evaluation value may be a numerical value that can index the evaluation tendency of the entire set from the conformity evaluation value of each article data in a predetermined set of article data.

FIG. 3 is a diagram showing a general display format of the web log.
Here, a general web log display format 170 is shown. The article data receiving unit 122 acquires article data shown in such a web log display format 170, and the range specifying unit 152 specifies unit data as a range to be evaluated. The first article data 182 is article data posted at the date and time indicated in the article date column 172 by the user A who has opened this web log. The article title column 174 is the title of article data. The article column 176 is the contents of article data. The first article data 182 includes an article date column 172, an article title column 174, and an article column 176.

  The second article data 184 is a comment posted from another user B on the article content of the first article data 182. The comment column 178 indicates the content of the comment, and the comment date column 180 indicates the posting date / time. The third article data 186 is article data posted by the user A at another date and time.

  As described above, the web log display format 170 includes a plurality of article data. The range specifying unit 152 specifies a range to be used as unit data using date information included in the article data as a key. The range specifying unit 152 uses the date format such as the article date column 172 and the comment date column 180 as a delimiter from the document data shown in the web log display format 170, and the web log display format 170 shown in FIG. It is recognized that three pieces of article data including data 182, second article data 184, and third article data 186 are included. And about the 1st article data 182, the data of the article title column 174 and the article column 176 are specified as unit data. For the second article data 184, the data in the comment field 178 is specified as unit data.

The information amount calculation unit 155 counts the number of characters in the unit data. For example, if the article title column 174 states that “Hanami is the best” and the article column 176 says “I went to Ueno Park for a group today”, the information amount calculation unit 155 will calculate the square root of the 27 characters. That is, 5.19 is calculated as the information amount of the unit data.
As another example, the information amount calculation unit 155 may set the title as a higher evaluation target than the content of the article. For example, the square root of (number of title characters × predetermined number + number of article characters) may be used as the information amount of the unit data. Alternatively, in the article column 176, a high score may be given to a sentence in which the target term is described in a special typeface such as a bold typeface. The information amount calculation unit 155 may specify unit data in the sentence column 176 in sentence units or paragraph units. At this time, the information amount calculation unit 155 calculates the information amount for unit data in sentence units and paragraph units.

  Note that when the document data does not include date information, a title, or other information for identifying the web log, the range specifying unit 152 excludes the document data from processing. That is, document data that does not match a predetermined web log display format, such as a news site or a company homepage, may be excluded from processing. Alternatively, the range specifying unit 152 may determine that the document data includes date information as a processing target. In this embodiment, the topic trend analysis is limited to the web log. However, as a modified example, the document classification process may be executed without being limited to the web log.

FIG. 4 is a schematic diagram when the topic trend is visually displayed.
The topic scale management apparatus 100 collects article data from the weblog site 300, specifies unit data, and calculates an information amount and a conformity evaluation value. The topic scale management apparatus 100 calculates a topic scale value and an overall evaluation value for a set of article data, for example, a set of article data classified into a certain category. The topic scale management apparatus 100 visually indicates a topic trend by displaying a graphic such as a graph reflecting the topic scale value and the overall evaluation value. FIG. 4 is an example of the display mode. Here, it is a screen display mode when illustrating the trend of topics related to five vehicle types from vehicle type A to vehicle type E.

  The set evaluation conditions are “whether the utterance is made by the purchaser” or “whether the utterance is made by an unpurchased person”. In the case of an article posted by a purchaser, terms such as “bought” and “favorite” are regarded as important terms for evaluation. In the case of an article posted by an unpurchased person, terms such as “want”, “want to buy”, and “how about” are regarded as important terms for evaluation. A vehicle type closer to the horizontal axis indicates a topic among non-purchasers. A vehicle type closer to the vertical axis indicates a topic among buyers. That is, this figure is expressed as (x, y) = (topic size between unpurchased people, topic size between purchasers). The farther away from the origin, the larger the topic scale. In the figure, the ratio between the number of articles posted by non-purchasers and the number of articles posted by buyers corresponds to a so-called overall evaluation value. That is, it can be considered that the distance from the origin reflects the topic scale value and the angle viewed from the origin reflects the overall evaluation value. In this case, it is not necessary to calculate the angle after the overall evaluation value is explicitly calculated, but it can be said that the angle is calculated as a value indicating the overall evaluation tendency, that is, the degree of conformity with the evaluation criterion. In this manner, in the case of an orthogonal coordinate system, it may be expressed as (x, y) = (topic size for one evaluation, topic size for another evaluation). Further, in the case of a polar coordinate system, it may be expressed as (r, θ) = (total topic size, value reflecting overall evaluation tendency). The direct coordinate expression and polar coordinate expression will be described in more detail with reference to FIGS. 12 and 13, respectively.

  In the figure, the model A is a topic among non-purchasers, but the topic scale is smaller than the model E. That is, it can be seen that the car model is not much talked about. The model B is a topic among purchasers, but the scale of the model is small compared to other models. On the other hand, it is presumed that the vehicle type C is a famous car that has a large topic scale among purchasers and is popular with owners. On the other hand, the vehicle type D has a large topic scale among non-purchasers, and is presumed to be a car model longed for by non-purchasers. It is presumed that the vehicle type E is a popular vehicle type that has no shortage of topics for both purchasers and non-purchasers.

In general, the life cycle of a product is a topic that is centered between unpurchased customers, but the level of market recognition is low and the topic scale is small, gradually becoming recognized in the market, and the topic scale between buyers increases, While the overall topic scale will increase, it will become a topic centered between buyers, but the topic scale itself will continue to shrink. The topic scale management apparatus 100 displays a graph using the topic scale value and the overall evaluation value as parameters for the article data set related to a specific topic in this way.
Next, a process for classifying highly related article data as a category will be described.

FIG. 5 is a schematic diagram for explaining categories.
In this embodiment, article data is classified into three categories. The first category group 190 indicates a first-stage category among them. As shown in the figure, the first category group 190 includes “entertainment”, “living / hobbies”, “travel / leisure”, “sports”, “computer / internet”, “business / economics”,. It consists of 10 categories shown as “Other”.

  The classification processing unit 159 calculates a category evaluation value indicating the degree of conformity with respect to each of the categories indicated in the first category group 190 for the acquired unit data. The conformity evaluation unit 156 classifies the article data into categories in which the category evaluation value is a predetermined value, for example, 80 or more. If there is no category of 80 or more, it is classified into the “other” category. When there are a plurality of categories of 80 or more, this article data is classified into the plurality of categories.

  The second category group 192 indicates a second-stage category obtained by further subdividing the “entertainment” category in the first category group 190. The category of the second category group 192 here can be said to be a lower category belonging to the “entertainment” category. As shown in the figure, the second category group 192 includes five categories indicated as “music”, “celebrity”, “outing”, “magazine”, and “others”. Article data belonging to the “entertainment” category is further classified into one of the categories shown in the second category group 192. Also at this time, the article data may be classified into a plurality of categories in the second category group 192.

  The third category group 194 indicates a third-stage category obtained by further subdividing the “music” category in the second category group 192. It can be said that the category of the third category group 194 here is a lower category belonging to the “entertainment” category or the “music” category. As shown in the figure, the third category group 194 includes seven categories indicated as “artist”, “genre”, “event / concert”, “instrument”, “karaoke”, “hit chart”, and “others”. Is done. Article data belonging to the “music” category is further classified into one of the categories shown in the third category group 194. Also at this time, the article data may be classified into a plurality of categories in the third category group 194.

The third category group 194 may be further subdivided into a fourth stage and a fifth stage. For example, “artist” may be further categorized by artist name. The category configuration shown in the figure is not necessarily fixed, and the user of the topic scale management apparatus 100 may redesign. At this time, division, addition, deletion, and integration of categories may be performed. Further, the data processing unit 130 may integrate a category with a small topic scale as an “other” category.
The set of categories need not be one type. Category composition is possible from various aspects. For example, the “music” category may be subdivided from the viewpoint of genre, such as “classic”, “rock”, “blues”, or “Japan”, “China”, “America”, “Nordic”. As such, it may be subdivided from a regional perspective. In such a case, the classification processing unit 159 may classify article data for each of a plurality of category sets. That is, according to the above-described example, the article data A is classified into the category “music> rock” in the category set 1 and the category “music> norse Europe” in the category set 2.

  Further, related words associated with the words may be defined for the words specific to the category. For example, various related terms such as the author, nationality, genre, singer name, and related works may be defined for a word indicating a specific song name. The classification processing unit 159 performs the category evaluation on the implied related terms in addition to the category evaluation based on the words specified in the article data, so that the category evaluation value of the article data is expressed and implied. May be calculated comprehensively.

FIG. 6 is a data structure diagram of the topic information holding unit.
The category column 200 corresponds to the three-stage category configuration shown in FIG. The absolute scale column 202 indicates the absolute topic scale value of each category. A relative scale column 204 indicates a relative topic scale value between categories. The overall evaluation column 206 shows the overall evaluation value for each category. This figure shows the topic scale information for the “entertainment” category of the first category group 190 in the topic scale information. Also, the figure shows topic scale information at a predetermined date and time, and the topic information holding unit 164 holds topic scale information for a predetermined period as a collection of topic scale information for each date and time.

  The topic scale value of each category is calculated as an added value of the information amount of the unit data classified into the category. In the figure, four pieces of unit data are posted in the “artist” category. Of these unit data, the first case is 100 characters, the second case is 100 characters, the third case is 100 characters, and the fourth case is 100 characters. Then, the information amount of each unit data from the first case to the fourth case is 10, which is the square root of 100. The topic scale value calculation unit 142 calculates the topic scale value of the “artist” category as 40 by 10 + 10 + 10 + 10 = 40. In the absolute scale column 202, the total number of article data and the total number of characters of each category are shown.

  On the other hand, in the “karaoke” category, three unit data are classified. At this time, it is assumed that the number of characters of the first to third article data is 10 characters, 30 characters, and 640 characters, respectively. In this case, by the same calculation, the respective information amounts are 3.16, 5.48, and 25.3. The topic scale value calculation unit 142 calculates the topic scale value of the “karaoke” category as a sum of 33.94, that is, approximately 34. The topic scale value calculation unit 142 adds the topic scale values of the categories belonging to the third category group 194 such as the “artist” category and the “karaoke” category, thereby adding the topic scale values of the “music” category, which is the higher rank. Is calculated as “363”.

Similarly, by adding the topic scale value of each category belonging to the second category group 192 such as the “music” category and the “celebrity” category shown in the second category group 192, the “entertainment” category corresponding to the upper category is added. The topic scale value is calculated as “3025”. Furthermore, the topic scale value “45987” as a whole is calculated by adding the topic scale values of each category belonging to the first category group 190 such as the “entertainment” category shown in the first category group 190.
In this way, hot topics on the Internet and topics that are not are quantified as absolute topic scale values based on the topic scale values calculated for each category.

  As described with reference to FIG. 5, the unit data may be classified into a plurality of categories. For example, it is assumed that the conformity evaluation value for “artist” of certain unit data is “95” and the conformity evaluation value for the “genre” category is “85”. Further, it is assumed that the information amount of the unit data is “360”. In this case, the information amount “323” of the article data is distributed at a ratio of 95:85 to the “artist” category and the “genre” category, respectively. That is, 360 × 95 / (95 + 85) = 190 is allocated to the “artist” category, and 360 × 85 / (95 + 85) = 170 is allocated to the “genre” category. Since one unit data may include a plurality of topics, the topic scale value of each category can be measured more accurately by such processing.

  The topic scale value calculation unit 142 also calculates a relative topic scale value. For example, the topic scale value for all article data collected by the topic scale management apparatus 100 is “45987”, of which the topic scale value in the “entertainment” category is “3025”. From 3025 ÷ 45987 = 0.0658, it is calculated that about 6.58% of the topics in the weblog are related to entertainment. Thus, the topic scale value calculation unit 142 calculates the relative topic scale value as a percentage expression.

  Since the topic scale value of the “music” category is “363”, it is calculated that approximately 0.79% of the topics in the web log are topics related to music by 363 ÷ 45987 = 0.79 (percent). The For each category in the third category group 194 belonging to the “music” category, relative topic scale values for all topics in the web log are calculated.

  Since the topic scale value of the “music” category is “363” with respect to the topic scale value “3025” of the “entertainment” category, 363 ÷ 3025 = 12.0 (percent), so that about the topic of entertainment related topics Twelve percent are calculated to be music topics. Similarly, for each category of the third category group 194 belonging to the “music” category, a relative topic scale value with respect to the topic scale value of the “entertainment” category is calculated.

Since the topic scale value of the “artist” category is “40” with respect to the topic scale value “363” of the “music” category, approximately 11% of the topics related to music from 40 ÷ 363 = 11 (percent). Is a topic about the artist. Similarly, for each category of the third category group 194 belonging to the “music” category, a relative topic scale value with respect to the topic scale value of the “music” category is calculated.
In this way, relative topic scale values are calculated for various categories.

  The overall evaluation value calculation unit 144 calculates an overall evaluation value for each category. For example, in the case of the evaluation condition of whether the article is a positive article or a negative article, the conformity evaluation unit 156 calculates the degree of being an affirmative article as the conformity evaluation value for each unit data. The overall evaluation column 206 shows the overall evaluation value for each category calculated in this way. In addition to the common conditions (hereinafter also referred to as “general evaluation conditions”), the evaluation conditions include conditions specific to the category (hereinafter also referred to as “attribute evaluation conditions”). For example, an automobile may be evaluated for each evaluation item such as “price” and “design”. For cosmetics, “price”, “moisturizing”, “goodness of dropping”, and the like can be considered as evaluation items. The topic information holding unit 164 may hold the overall evaluation value for each evaluation item according to the category. The overall evaluation column 206 in the figure shows the overall evaluation value only for one general evaluation condition of whether the content is affirmative, but the topic information holding unit 164 further has empirical content or speculative content. You may hold | maintain the whole evaluation value about several general evaluation conditions, such as whether there exists. The topic information holding unit 164 may also hold a plurality of overall evaluation values for attribute evaluation conditions. Moreover, the trend of the topic can be analyzed from the conformity evaluation values for the general evaluation condition and the attribute evaluation condition. For example, if it is specified that the topic is “moisturizing” of “cosmetics A” as the first viewpoint and the topic is “positive” and “experienced” as the second viewpoint, It is possible to determine that “A is popular with purchasers in terms of moisture retention”.

FIG. 7 is a flowchart showing a process of categorizing article data.
First, the article data receiving unit 122 receives article data from the web log site 300 (S10). The range specifying unit 152 specifies unit data that is a range to be evaluated from the acquired article data (S12). The information amount calculation unit 155 calculates the information amount of the unit data (S14). The conformity evaluation unit 156 calculates conformity evaluation values of the unit data according to the set evaluation conditions (S16). When a plurality of types of evaluation conditions are set, the conformity evaluation unit 156 calculates a plurality of types of conformity evaluation values corresponding to each evaluation condition. The classification processing unit 159 calculates the category evaluation value for the contents of the unit data, and specifies the category into which the unit data should be classified (S18).

When there are two or more categories with a category evaluation value of 80 or more (Y in S20), the information amount calculated in S14 is distributed according to the category evaluation value (S22). When the category having a category evaluation value of 80 or more is 1 or less (N in S20), the process in S22 is skipped. At this time, the unit data is classified into one of the categories or the “other” category. The topic scale value calculation unit 142 recalculates the absolute topic scale value from the information amount of the unit data (S26). At this time, the topic scale value calculation unit 142 similarly recalculates the relative topic scale value. The overall evaluation value calculation unit 144 recalculates the overall evaluation value for each category (S28). Thus, the topic scale information of the topic information holding unit 164 is updated.
Here, the description has been made on the assumption that the article data is weblog data. In S12, if the received article data does not match the predetermined display format, the range specifying unit 152 may be excluded from the evaluation target and not execute the subsequent processing.

FIG. 8 is a flowchart showing a processing process when investigating a topic scale trend based on a category.
The topic scale management apparatus 100 can graphically display the topic scale and the topic tendency for each category in the manner shown in FIG. At this time, the user sets the condition of the article data to be investigated. In the present embodiment, in the topic scale trend analysis based on the category, the following four types of settings are assumed as the setting method of the survey target.
(1) Specify a plurality of categories and compare topic trends among the specified categories (hereinafter referred to as “first setting”).
For example, based on the example of FIG. 5, three categories of “entertainment>music> artist”, “entertainment>music> genre”, and “entertainment> magazine” may be explicitly specified as survey targets.
(2) Specify one category explicitly or implicitly, and compare topic trends between lower categories of the designated category (hereinafter referred to as “second setting”).
For example, when one category is designated as “entertainment”> music ”,“ artist ”,“ genre ”,“ event concert ”,“ instrument ”,“ karaoke ” Each category of “Hit Chart” and “Other” is the subject of investigation. In addition, when the category is not explicitly designated, each category of “entertainment”, “living / hobbies”,..., “Politics”, “others” of the first category group 190 is the subject of the survey. .
(3) A plurality of attribute evaluation conditions for one category are specified, and topic trends are compared for each evaluation item of the specified category (hereinafter referred to as “third setting”).
For example, for the category “cosmetics”, evaluation items such as “price” and “humidity retention” are set in advance as attribute evaluation conditions for cosmetic evaluation. If the category “cosmetics” and this “attribute evaluation condition for cosmetics” are designated, the topic scale and topic trend can be displayed for each evaluation item.
(4) One category is specified explicitly or implicitly, and the topic trend of the specified category is displayed (hereinafter referred to as “fourth setting”).
This is a case where comparison display is not performed. For example, when one category is designated as “entertainment”> music ”, only the topic scale and topic trend in that category are investigated. In addition, when a category is not explicitly specified, the topic scales and topic trends of all categories are investigated.

  When the user designates a category name as an investigation target (Y in S30) and a plurality of category names are designated (Y in S32), this corresponds to the first setting. The topic scale value calculation unit 142 calculates the topic scale value for each designated category, and the overall evaluation value calculation unit 144 also calculates the overall evaluation value for each designated category. In this way, topic trends for each category are tabulated as topic scale information (S34). The evaluation condition may be set in advance in the article data collection stage. In this case, the overall evaluation value can be acquired by referring to the overall evaluation column 206 of the topic information holding unit 164 during the survey. On the other hand, the evaluation condition may be set at the time of the survey. In this case, the overall evaluation value is calculated by evaluating the article data in the article data holding unit 162 at the time of the survey. The same applies to the second to fourth settings.

  On the other hand, when the user does not specify a category name (N in S30) or when the specified category is single (N in S32), the investigation determination unit 153 is set to the second setting, that is, Then, it is determined whether or not the investigation condition is set so that the lower category should be searched (S36). When the lower category is to be investigated (Y in S36), the statistical unit 140 aggregates topic trends for each of the lower categories as topic scale information (S38).

In the third setting, that is, when the survey condition is set so that topic trends should be investigated for a plurality of evaluation items for the specified category (Y in S40), the statistical unit 140 determines each evaluation item of the specified category. The topic trend about is aggregated as topic scale information (S42). Further, if the survey is not for a plurality of evaluation items (N in S40), only the topic trend for the category designated as the fourth setting is the subject of the tabulation process (S44).
Thus, the topic trends aggregated in any of S34, S38, S42 and S44 are displayed on the screen (S46).

As shown in FIG. 5 and the like, the topic of the article data may be classified into any category that has been categorized in advance. For example, when comparing the topic trends of the vehicle type A and the vehicle type B, the user of the topic scale management device 100 may create a category corresponding to the vehicle type A or the vehicle type B. The user can flexibly set the survey target range by freely typifying a set of article data to be compared.
As another method, by specifying a set of article data including the word “car type A” and specifying a set of article data including the word “car type B”, topic trends in the respective sets may be compared. In this case, since the user can classify the set of article data simply by specifying a word, the survey target range can be set more easily. Hereinafter, a set of article data including the word A is referred to as “set related to A”. In FIG. 8, the second setting for investigating the lower categories of a certain category has been described. As a similar aspect of this setting, a set of article data including a word related to a specified word may be a survey target. First, the principle and purpose of the related term search in this embodiment will be described with reference to FIG. 9, and then the topic trend analysis processing based on the term will be described with reference to FIG.

FIG. 9 is a schematic diagram showing the concept of related term search.
The topic scale management apparatus 100 according to the present embodiment has a function of detecting a term highly relevant to a search target term based on a topic trend. Here, the purpose and principle will be described by taking the case where the term “drive” is a designated term as an example.
The first document group 208 is a set related to “drive”. The area of the first document group 208 indicates the topic scale value. The second document group 210 is a set relating to “stress”. A portion where the first document group 208 and the second document group 210 overlap, that is, a product set of the first document group 208 and the second document group 210 is a set of article data including both the terms “drive” and “stress”. Become. Now, when the first document group 208 is specified using the term “drive” as a search target term, it is assumed that the most frequently used term is “stress” except for “drive”. From such information, we will hypothesize that “drive” is generally an effective technique for relieving “stress”.

Next, when the second document group 210 is specified using the term “stress” as a search target term, it is assumed that the most frequently used term is “game” except for “stress”. Then, it is possible to speculate that “game” is also a powerful means as a “stress” elimination method. In other words, both “drive” and “game” may be in a competitive relationship in terms of how to relieve stress. Conceptually, “drive” and “game” are not so closely related terms, but due to the common topic of “stress”, these seemingly unrelated terms are unexpectedly A close relationship will be discovered.
In summary, the second term “stress” is identified from the first term “drive”, and further the third term “game” is identified, whereby the relationship between the first term “drive” and the third term “game” is determined. It is possible to find sex. The topic scale management apparatus 100 in the present embodiment detects related terms based on topic trends from specific terms based on such a principle.

  Here, “game” is detected by automatically setting “stress”, which appears frequently after “drive” in the first document group 208, as a search target term, but the second term is not necessarily the second term. It may not be the word that appears next to “drive” in the document group 210. For example, with the first term “drive” and the second term “accident”, it is possible to find a third term associated with the first term “drive” with respect to “accident”.

In this embodiment, the second term is referred to as the “subordinate term” of the first term for convenience. The third term is referred to as the first term “equivalent term”.
By comparing the first document group 208 and the third document group 212 that is a set related to the third term “game”, the topic scales for the equivalence terms can be compared. Further, by comparing the size of the product set of the first document group 208 and the second document group 210 with the size of the product set of the second document group 210 and the third document group 212, the first term and the second term relating to the second term The degree of relevance of the three terms can also be analyzed. As described above, the topic scale management apparatus 100 according to the present embodiment supports the discovery of the relationship between terms based on the topic trend after analyzing the topic trend using the term as a key.

FIG. 10 is a flowchart showing a processing process when investigating topic scale trends based on terms.
In the present embodiment, in the topic scale trend analysis based on the terms, the following five settings are assumed as the setting method of the survey target.
(5) Designate a plurality of terms and compare topic trends between sets related to each designated term (hereinafter referred to as “fifth setting”).
For example, three terms “drive”, “game”, and “music” may be explicitly specified as the investigation target.
(6) Specify one term and compare topic trends between sets related to each subordinate term of the designated term (hereinafter referred to as “sixth setting”).
For example, when one term is designated as “drive”, the term detection unit 158 detects three terms “stress”, “car”, and “hobby” as subordinate terms. At this time, a set related to these terms is a subject of investigation.
(7) Specify one term and compare topic trends between sets for each coordinated term of the specified term (hereinafter referred to as “seventh setting”).
For example, when one term is designated as “drive”, the term detection unit 158 detects three terms “game”, “tennis”, and “karuta” as its equivalence terms. At this time, a set related to these terms is a subject of investigation.
(8) A plurality of attribute evaluation conditions for one term are specified, and topic trends are compared for each evaluation item of the specified term (hereinafter referred to as “eighth setting”).
For example, for the term “cosmetic A”, evaluation items such as “price” and “humidity retention” are set in advance as attribute evaluation conditions for cosmetic evaluation. If the term “cosmetic A” and this “attribute evaluation condition for cosmetics” are designated, the topic scale and the topic tendency can be displayed for each evaluation item.
(9) One term is explicitly specified, and the topic trend of the set related to the specified term is displayed (hereinafter referred to as “the ninth setting”).
For example, when one term is designated as “drive”, only the topic scale and topic trend in the set relating to the “drive” are to be investigated.

  When a plurality of terms are designated as a survey target (Y in S50), this corresponds to the fifth setting. The topic scale value calculation unit 142 calculates the topic scale value for each set related to each designated term, and the overall evaluation value calculation unit 144 also calculates the overall evaluation value for each set related to each designated term. Thus, topic trends for each set relating to each term are aggregated as topic scale information (S52). At this time, the evaluation condition may be set in advance or may be specified at the time of the survey process. The same applies to the sixth to ninth settings.

  On the other hand, when the designated term is single (N in S50), the investigation determination unit 153 determines whether or not the investigation condition is set to search for the sixth setting, that is, the lower term (S54). When subordinate terms are to be investigated (Y in S54), the term detection unit 158 lists subordinate terms that frequently appear in the set relating to the designated term (S56). The statistics unit 140 aggregates topic trends for each of the sets related to the lower-level terms as topic scale information (S58).

  When the designated term is single (N in S50) and the investigation determination unit 153 is in the seventh setting, that is, when the investigation condition is set to search for the equivalent term (Y in S60), the term The detecting unit 158 lists the subordinate terms that frequently appear in the designated term set (S62). Of these, the term detection unit 158 selects a subordinate term that appears most frequently other than the designated term (S64). By the method shown in FIG. 9, the equivalence term of the designated term is listed from the selected subordinate terms (S66). The statistic unit 140 aggregates topic trends for each set of equivalence terms as topic scale information (S68).

In the eighth setting, that is, when the survey condition is set as the topic term should be investigated for a plurality of evaluation items for the designated term (Y in S70), the statistical unit 140 sets the evaluation term for each evaluation item in the set relating to the designated term. Topic trends are tabulated as topic scale information (S72). If the survey is not for a plurality of evaluation items (N in S70), that is, in the case of the ninth setting, the topic trend for the designated term is the subject of the aggregation process (S74).
Thus, the topic trends aggregated in any of S52, S58, S68, S72 and S74 are displayed on the screen (S76).

The user may select at the time of the survey whether to identify the survey target based on the category or to identify the survey target based on the term. Alternatively, if a term designated as a survey target by the user matches the category name, the survey target may be specified based on the category, and if not matched, the survey target may be specified based on the term.
Next, various topic trend display modes will be described as examples other than FIG.

FIG. 11 is a diagram showing topic scale information in a plurality of topics in a table format.
The figure shows the topic trend analysis results for four vehicle types from vehicle type A to vehicle type D. Here, it is assumed that the article data is divided into two depending on whether it is a positive article. In other words, the conformity evaluation value is binarized in response to positive or negative. The vehicle type A to the vehicle type D may be category names or terms.
As for the vehicle type A, the positive article is 73 and the negative article is 41. Based on the information amount of each article, the topic scale value for the vehicle type A is calculated. According to the figure, it can be seen that there are many positive articles for the vehicle type A, but there are many negative articles for the vehicle type D. In addition, the number of posted articles is small for model C. A display mode when such data is obtained is shown in FIGS.

FIG. 12 is an example of a graph showing the topic scale information of FIG.
The scale and trend of topics from the vehicle type A to the vehicle type D as shown in FIG. 11 are expressed in an orthogonal coordinate system. In the figure, (x, y) = (number of negative articles, number of positive articles). As in FIG. 4, it can be considered that the distance from the origin reflects the size of the topic scale value, and the angle reflects the overall evaluation value as the overall evaluation tendency. Here, the overall evaluation value = the number of positive articles ÷ the number of negative articles may be used. The display unit 114 displays the topic trend for each topic to be compared in S46 in FIG. 8 or S76 in FIG.

FIG. 13 is another example of a graph showing the topic scale information of FIG.
The topic scales and trends from the vehicle type A to the vehicle type D as shown in FIG. 11 are represented in the polar coordinate system in the figure. When m = the number of positive articles and n = the number of negative articles, the distance r = m + n and the angle θ = 45 degrees × (mn) / (m + n). Similar to FIG. 12, the distance from the origin reflects the size of the topic scale value, and the angle reflects the overall evaluation value as the overall evaluation tendency. The topic trend for each topic is also illustrated by such a screen display mode.

FIG. 14 is a graph showing the temporal transition of the topic trend.
Here, in the Cartesian coordinate system as shown in FIG. 12, the trend of the topic trend of each vehicle type is shown. As described with reference to FIG. 4, the scale and trend of the topic change with time. FIG. 14 shows changes in topic trends during the specified period. For example, in the case of the vehicle type A, the topic scale is steadily expanding with time. Moreover, it turns out that the content of the article is also generally affirmative. In the case of the vehicle type B, the scale and trend of the topic hardly change with time. From this, it can be seen that the vehicle type B is a vehicle type whose evaluation is stable. In the case of vehicle type D, the topic content changes from a positive tendency to a negative tendency with the passage of time. That is, in the case of the vehicle type D, it can be seen that the evaluation deteriorates with time.
According to such a display mode, a temporal change in the topic trend of each vehicle type can be grasped visually and intuitively. Note that the user can also display a part or all of the contents of the article data that is the object of evaluation at that time by clicking an arbitrary point on the topic transition line for each vehicle type with the mouse.

FIG. 15 is a schematic diagram of a model that is widely used to show the market evaluation of products.
Portfolio models are widely used in marketing strategies to show market positions for specific products. The topic scale management apparatus 100 in the present embodiment can also display a model similar to this portfolio model based on the topic scale information. Before that, a portfolio model that is generally used and is based on this embodiment will be briefly described.

The model shown in the figure is a model for illustrating the position of each product in the market for a product group having a competitive relationship in a certain market. The horizontal axis shows the market share of the surveyed products. The left side shows a higher share and the right side shows a lower share. The vertical axis shows the growth rate of the market as a whole. The upper side indicates a growth market with a higher growth rate, and the lower side indicates a mature market with a slower growth rate. The area shown in the figure can be divided into four areas.
(1) Money Trees This area includes products that have a high share in mature markets. Shows that it is a standard product that generates cash flow without significant additional investment.
(2) Stars This area includes products that have a high market share in growth markets. Shows that it is a market leader and an expected product that increases cash flow.
(3) Problem children In this area, products that do not have a share in the growing market belong. Although it is promising as a market, it shows that it is a product that has lost opportunity because it has not gained market share.
(4) Losing dogs In this area, products that do not have a share in the mature market belong. It is not attractive as a market, and it is a product that has become a baggage because it has not been shared.

  Depending on which of the areas shown in the figure the surveyed product belongs, it can be used to determine the marketing strategy for that product. The portfolio model shown in the figure is based on real numbers of actual market growth rate and share of products actually put on the market. The present inventor has come up with the idea that a similar model can be constructed not by actual numerical values such as the market scale but by a market leading indicator of the topic scale.

FIG. 16 is a model diagram for illustrating product evaluation based on the topic scale.
Here, a hypothetical market composed of three vehicle types, vehicle type A, vehicle type B, and vehicle type C, is assumed. These vehicles are assumed to be competitive products in the market. The horizontal axis shows the topic scale share of the surveyed products. For example, in the market at a certain date and time,
Topic scale value of car type A / (Topic scale value of car type A + Topic scale value of car model B + Topic scale value of car model C)
It is calculated by. The left side shows a higher share and the right side shows a lower share.
The vertical axis shows the topic scale growth rate of the market as a whole. The upper side shows that the growth rate is larger, and the lower side shows that the growth rate is slowing down. The topic scale growth rate is obtained from the ratio of the topic scale value of the entire market at the survey target date and time t to the topic scale value at the date and time t-1 one day before. That is,
Topic scale growth rate = (topic size value of market at date and time t) ÷ (topic size value of market at date and time t−1) −1
It is calculated by. Needless to say, the topic size value of the market is obtained by the topic size value of the vehicle type A + the topic size value of the vehicle type B + the topic size value of the vehicle type C.

  In the figure, at the stage of January 1, among the vehicle types A, B, and C, the topic share regarding the vehicle type A is large, and the shares of the vehicle type B and the vehicle type C are small. On the 1st of February, the topic scale growth rate is increasing. Moreover, the share of the vehicle type B and the vehicle type C has expanded, and the share of the vehicle type A has decreased correspondingly. However, this does not necessarily mean that the topic scale of the vehicle type A has decreased, but is a relative value to the topic scale values of the vehicle type B and the vehicle type C. At the stage of March 1, the topic scale growth rate is even higher. The share of model B has increased significantly. That is, the vehicle type B has entered the star area shown in FIG. On the other hand, vehicle type A has lost its market share and has entered the area of problem children. Car type C does not gain as much market share as car type B, and is still in the area of problem children.

  At the April 1st stage, the topic scale growth rate is the highest. The vehicle type B remains in the star region, and the vehicle types A and C remain in the problem child region. On May 1st, topical growth slowed down, and on June 1st, it further slowed down. During this period, the share of the vehicle type B has fallen, but has entered the area of money trees. On the other hand, vehicle type A and vehicle type C are entering the area of losers. Finally, on July 1st, the topic scale growth rate has further declined.

  In this way, the display unit 114 can display the topic scale and trend transition of the vehicle types A, B, and C in a predetermined period so as to be comparable on the same screen. The topic scale reflects the interest of contributors and is considered to be effective as a leading indicator for various indicators such as purchasing activities and brand image. According to such an aspect, the evaluation between competing products can be quantified based on the topic, and data can be presented easily for the user to visually recognize.

FIG. 17 shows a screen that displays topic trends for each attribute evaluation condition.
Finally, an example of the screen regarding the data display mode in the third setting and the eighth setting is shown. The evaluation criteria vary depending on the evaluation target. For example, the evaluation criteria for automobiles are price, design, ride comfort, and performance. Even a car model that is positively evaluated as a whole may be negatively evaluated in terms of design. On the other hand, there may be a car model that has a low overall evaluation but a high design. The topic scale management device 100 can also illustrate the topic trend for each evaluation item for a plurality of vehicle types A, B, and C.

In the figure, the pie chart shown in the upper half shows topic scale information about favorable article data. The size of the area of the pie chart reflects the size of the topic scale value. For example, the radius of the pie chart may be defined by the topic scale value. The pie chart shown in the lower half shows topic scale information about unsatisfactory article data.
According to the figure, in the case of the vehicle type A, the topic scale of the unsatisfactory article is larger than the favorable article. On the other hand, the model B has a larger topic size of favorable articles than unsatisfactory articles. Car model C has a small topic size for both favorable and unsatisfactory articles. That is, it can be seen that this is a car model that has not been much talked about. It can be seen that the vehicle type A is favorably evaluated for the design, and the vehicle type C is unsatisfactory in terms of price. According to such a display mode, the evaluation between competing products can be further subdivided and analyzed. Note that part or all of the article data may be displayed by clicking the pie chart with the mouse. For example, when the user h clicks the “performance” area in the pie chart of the favorable article of the vehicle type B, the corresponding article data that is the basis of the topic trend totaling process may be displayed on the screen.

  As described above, according to the topic scale management apparatus 100 in the present embodiment, it is possible to collect information from personal media such as the weblog site 300 and analyze the topic trend of the world.

  In order to collect personal opinions through a questionnaire via the Internet, it is necessary to first make the user aware of the existence of the questionnaire. For this purpose, a device for attracting users to the questionnaire homepage is required. Moreover, in order to stimulate the questionnaire response will be devised to give a sweepstakes, but it is also true that such a method is costly. Furthermore, since it is often necessary to transmit personal information when answering a questionnaire, and the user tends to feel psychological resistance, this is an obstacle to the willingness to answer the questionnaire.

  Some news sites have a mechanism for polling specific current affairs, but they also have similar problems. In addition, since the questionnaire collects data as an answer to the question, in other words, there is a demerit that it is difficult to collect incidental information that deviates from the question. Furthermore, there are cases where half of the answers are made to the questionnaire, and noise information such as this may deteriorate the reliability of the questionnaire.

  Other systems that collect personal opinions include methods such as audience rating surveys and site access count surveys. In such a survey method, it can be said that problems such as a user's psychological resistance and labor are alleviated as compared with a questionnaire, but the information obtained is extremely limited.

  On the other hand, the topic scale management apparatus 100 of the present embodiment analyzes the trend of the world by collecting web log data in which individual opinions are expressed. The main benefits are as follows: 1. Ingenuity to stimulate the respondents' willingness to respond is virtually unnecessary. 2. For survey subjects such as weblog openers, there is no need to feel that they are cooperating in surveys. 3. Since it is not necessary to put a user's opinion into a model according to a question item, various topic trends can be collected. 4). Since the topic scale management device 100 only accesses an existing weblog site, it has a high affinity with an existing system. 5. It is considered to be statistically stable because a large amount of information on the Internet is used as an information source. 6). Time series data of topic transitions can be easily collected. 7). It is possible to quantify vague information that is highly useful in terms of absolute scale and relative scale. 8). Relevance between terms viewed from topic trends can be detected. 9. You can specify any comparison target and visually compare their topic trends.

In marketing, etc., it is important to accurately catch product preferences and market reactions. The topic scale management apparatus 100 according to the present embodiment can be applied to various analyzes such as advertising cost effectiveness and discovery of a new market. For example, if it is a mass media person, it is necessary to consider what kind of news people are interested in and what kind of information they are looking for, and then consider the amount of information to be provided. For example, even if it is big news, if you get bored of the world, you should actively provide other news. On the other hand, news that the world is interested in over time should be actively provided. In addition, it is possible to detect what kind of information is related to a certain news by using related information search based on topic trends, which is effective for finding potential information needs. Tool. In addition, by visually displaying the relationships and transitions seen from the topic trends of a plurality of topics, it becomes easier to intuitively recognize competitive relationships in a virtual market based on topics.
For those involved in the manufacturing industry, a method may be considered in which the product is reflected in the production plan depending on how much the released product is talked about and how the topic scale is changing. As described with reference to FIG. 9, it is also possible to find out what a product is, or what kind of thing is highly topic-relevant with the product name as a search target term. Detecting relationships between terms as viewed from topic trends rather than conceptually fixed term structures is an effective tool for finding potential competitors and potential markets in business planning. By comparing a set of article data categorized in advance as a category or a set of article data specified in a termbase, the position of the survey target based on the topic scale can be clearly indicated.
As described above, the topic scale management device 100 shown in the present embodiment is effective in bringing the service provider and the receiver closer together.

On the other hand, since weblogs are being opened one after another, there is no guarantee that the opened weblogs can actually be seen by others. In fact, it's just a “self-speaking” site with many weblogs connected to the Internet. The topic scale management apparatus 100 automatically collects and analyzes the opinions of such web logs. For this reason, there is an advantage that weblogs can easily reflect their opinions in society. As described above, the topic scale management device 100 has an effect of inviting an individual to transmit information.
It can be said that the topic scale management apparatus 100 contributes to realizing an activated society in which individuals actively transmit their opinions.

It should be noted that the topic scale management device 100 is effective for marketing and public opinion surveys, but is also effective for general users to know the trend of public topics. In the present embodiment, the web log is used as an information source. However, the document data collected by the topic scale management apparatus 100 may be various, such as a news site on the Internet.
Moreover, the topic scale management apparatus 100 may expect the accuracy of the survey by excluding article data that impairs the reliability of data analysis such that a specific term is meaninglessly linked to each other.

  The present invention has been described above based on the embodiment. The present invention is not limited to this embodiment, and various modifications thereof are also effective as aspects of the present invention.

The functions of the topic trend display unit and the transition display unit described in the claims are mainly realized by the display unit 114 in this embodiment. The function of the investigation determination unit described in the claims is mainly realized by the investigation determination unit 153 and the term detection unit 158 in this embodiment.
It should also be understood by those skilled in the art that the functions to be fulfilled by the constituent elements described in the claims are realized by the individual functional blocks shown in the present embodiment or their linkage.

  In this embodiment, the feature vector method is used to calculate the conformity evaluation value and the category evaluation value. However, other known natural language processing methods such as the Bayesian filter method can be used. It will be understood by those skilled in the art.

It is a hardware block diagram of a topic scale management system. It is a functional block diagram of a topic scale management device. It is a figure which shows the general display format of a web log. It is a schematic diagram when displaying a topic trend visually. It is a schematic diagram for demonstrating a category. It is a data structure figure of a topic information holding part. It is a flowchart which shows the process which classifies article data. It is a flowchart which shows the processing process when investigating a topic scale trend based on a category. It is a schematic diagram which shows the concept of a related term search. It is a flowchart which shows the processing process when investigating a topic scale trend based on a term. It is a figure which shows the topic scale information in a some topic in a table format. It is an example of the graph which shows the topic scale information of FIG. It is another example of the graph figure which shows the topic scale information of FIG. It is a graph which shows the time transition of a topic trend. FIG. 2 is a schematic diagram of a model that is widely used to show market evaluation of products. It is a model figure for showing evaluation of a product based on a topic scale. It is a figure which shows the screen which displays the topic trend for every attribute evaluation conditions.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Topic scale management system, 100 Topic scale management apparatus, 110 User interface processing part, 112 Input part, 114 Display part, 120 Communication processing part, 122 Article data receiving part, 130 Data processing part, 140 Statistics part, 142 Topic scale value Calculation unit, 144 Overall evaluation value calculation unit, 146 Evaluation standard setting unit, 150 Evaluation unit, 152 Range specifying unit, 153 Investigation determination unit, 154 Investigation condition setting unit, 155 Information amount calculation unit, 156 Conformity evaluation unit, 158 Term detection 159, classification processing unit, 160 data storage unit, 162 article data holding unit, 164 topic information holding unit, 170 web log display format, 300 web log site.

Claims (12)

A document data receiving unit for receiving document data from an external device;
An investigation condition setting unit for setting an investigation condition for identifying document data to be investigated by an instruction input from the user;
A survey determination unit that refers to the survey condition and determines whether or not the received document data is to be surveyed;
An information amount calculation unit for calculating the information amount of the document data;
A topic scale value calculation unit for calculating, as a topic scale value, a cumulative information amount for a survey document set, which is a set of document data set as a survey target, from an information amount of document data set as a survey target;
A conformity evaluation unit that calculates the conformity evaluation value of received document data by using an evaluation function for indexing the degree of conformance of the document contents with the evaluation criteria defined for evaluating the content of document data as conformity evaluation values. When,
An overall evaluation unit that specifies, as an overall evaluation value, a degree of conformity to the evaluation standard as the survey document set from the conformity evaluation value of the document data that is the subject of the investigation
A topic trend display unit that visually displays the scale and trend of the topic in the survey document set, with the graphic formed using the overall evaluation value and the topic scale value as parameters,
With
The evaluation criterion is a criterion for determining whether the content of the document data is a statement made by a purchaser or a statement made by a non-purchaser,
2. The topic scale management apparatus according to claim 1, wherein the overall evaluation value is a ratio of the number of articles posted by a purchaser and the number of articles posted by a non-purchaser .   The topic trend display unit visually displays a topic scale and a trend in the survey document set by a vector having one of the topic scale value and the overall evaluation value as a direction and the other as a direction. The topic scale management device according to claim 1.   The topic scale management apparatus according to claim 1, wherein the evaluation criterion is a criterion for determining whether the content of the document data is a positive content or a negative content. When the survey determination unit identifies a plurality of types of survey document sets based on the survey conditions,
4. The topic scale according to claim 1, wherein the topic trend display unit displays a topic scale and a trend for each set of survey documents so that they can be compared on the same screen. 5. Management device. A classification processing unit for classifying the received document data into any of a plurality of categories hierarchized according to the content thereof;
When the survey condition setting unit sets the survey condition as document data to be classified as a category specified by the user,
The topic according to any one of claims 1 to 4, wherein the survey determination unit specifies a plurality of types of survey document sets for each of a plurality of subcategories obtained by further subdividing the designated category. Scale management device. When the survey condition setting unit sets the survey condition as to be subject to document data including any of a plurality of types of terms designated by the user,
The topic scale management device according to claim 1, wherein the survey determination unit specifies a plurality of types of survey document sets corresponding to each of the plurality of types of terms. When the survey condition setting unit sets the survey condition as to investigate document data including any of a plurality of types of subordinate terms for the term specified by the user,
The survey determination unit detects the plurality of types by detecting a plurality of types of terms that are included in a set of document data including the specified terms and that are different from the specified terms as subordinate terms. The topic scale management apparatus according to claim 1, wherein a plurality of types of survey document sets are specified corresponding to each of the subordinate terms. When the survey condition setting unit sets the survey condition as a search target for document data including any of a plurality of types of equivalence terms for a term specified by a user,
The survey determination unit detects a term different from the specified term as a subordinate term in a set of document data including the designated term, and detects the terminology of the document data including the subordinate term. A plurality of types corresponding to the plurality of types of equivalence terms are detected by detecting a plurality of types of terms different from the specified term and the subordinate terms as included in the set. The topic scale management device according to claim 1, wherein a survey document set is specified. When multiple types of evaluation criteria are established,
The conformity evaluation unit calculates a plurality of types of conformity evaluation values of the received document data corresponding to each evaluation criterion,
The overall evaluation unit identifies a plurality of types of overall evaluation values as an overall evaluation in the survey document set corresponding to each evaluation criterion,
The topic trend display unit visually displays a topic scale and a trend in the survey document set by a graphic formed using the topic scale value and the plurality of types of overall evaluation values as parameters. Item 10. The topic scale management device according to any one of Items 1 to 8.   It further comprises a transition display unit that refers to date information included in the document data and displays on the screen a transition of one or both of the topic scale value and the overall evaluation value calculated based on the document data in a predetermined period. The topic scale management device according to any one of claims 1 to 9. When the survey determination unit identifies a plurality of types of survey document sets based on the survey conditions,
The transition display unit includes a transition of an increase rate of a topic scale value as a whole of the plurality of types of survey document sets, and a transition of a ratio of the topic scale value of each survey document set to the topic scale values of the plurality of types of survey document sets as a whole. The topic scale management apparatus according to claim 10, wherein the topic scale is displayed on a screen.   The topic scale management apparatus according to claim 1, wherein the document data receiving unit receives an article electronically posted as a web log as the document data.

Images