JUDGMENT PROGRAM, JUDGMENT METHOD, AND INFORMATION PROCESSING APPARATUS

Abstract

An information processing device according to the present invention determines the trends of changes in index values that respectively indicate the degrees of a plurality of positive emotions of a plurality of users on the basis of the opinions of the plurality of users on a specific topic and / or on the basis of information pertaining to agreement / disagreement to the opinions of others on the specific topic. When the information processing device detects, among the trends of the changes in the index values that indicate the degrees of the plurality of positive emotions, a first trend and a second trend for which the difference therebetween is equal to or less than a threshold, the information processing device outputs, as users included in a target group, a first user corresponding to the first trend and a second user corresponding to the second trend among the plurality of users.

Description

[Technical field]

[0001] The present invention relates to a determination program. [Background technology]

[0002] Majorities and minorities may occur in various situations, such as discussions in communities such as SNS (Social Networking Service), discussions in web conferences, or places where opinions are posted about a certain topic or theme. Minorities and majorities are identified using analytical methods based on surveys, and in recent years, negative / positive analysis of opinions on a specific topic has become known. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Daiya Takamura, Takashi Inui, Manabu Okumura, "Extracting Word Sentiment Polarity Using Spin Model", Transactions of Information Processing Society of Japan, Vol. 47 No. 02 pp. 627-637, 2006. Summary of the Invention [Problem to be solved by the invention]

[0004] However, minority users tend to hide or weaken their opinions. Also, there are variations in the same positive or negative opinions. For this reason, a technology that simply judges the positive or negative opinions of a particular topic cannot capture the true opinions of users. Therefore, the above technology may not be able to identify users who have similar opinions on a particular topic from among multiple users.

[0005] In one aspect, an object of the present invention is to provide a determination program, a determination method, and an information processing device that are capable of identifying users who have similar opinions. [Means for solving the problem]

[0006] In a first proposal, the judgment program is characterized in that it causes a computer to execute a process to judge the trend of change in index values ​​indicating the degree of multiple positive emotions of multiple users based on at least one of the users' own opinions on a specific topic and information regarding the approval or disapproval of other people's opinions on the specific topic, and when a first trend and a second trend, in which the difference between the trends of change in index values ​​indicating the degree of the multiple positive emotions is detected and the difference between the first trend and the second trend is below a threshold, output a first user of the multiple users corresponding to the first trend and a second user corresponding to the second trend as users included in a target group. Effect of the Invention

[0007] According to one embodiment, users with similar opinions may be identified. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a diagram illustrating an example of the overall configuration of a system according to a first embodiment. [Diagram 2] FIG. 2 is a diagram for explaining extraction of opinion changes according to the first embodiment. [Diagram 3] FIG. 3 is a functional block diagram of the information processing device according to the first embodiment. [Figure 4] FIG. 4 is a diagram showing an example of user information stored in the user information DB. [Diagram 5] FIG. 5 is a diagram illustrating an example of information stored in the keyword DB. [Figure 6] FIG. 6 is a diagram illustrating an example of information stored in the posted information DB. [Figure 7] FIG. 7 illustrates a strong entrainment signal. [Figure 8] FIG. 8 illustrates the detection of a strong synchronization signal. [Figure 9]FIG. 9 is a diagram illustrating method 1 for determining a strong synchronization signal. [Figure 10] FIG. 10 is a diagram illustrating a second method for determining a strong synchronization signal. [Figure 11] FIG. 11 illustrates a weak entrainment signal. [Figure 12] FIG. 12 is a diagram illustrating a method for determining a weak synchronization signal. [Figure 13] FIG. 13 is a diagram illustrating detection of synchronized users. [Figure 14] FIG. 14 is a diagram illustrating detection of a protected user. [Figure 15] FIG. 15 is a flowchart of the process according to the first embodiment. [Figure 16] FIG. 16 is a diagram illustrating an example of a hardware configuration. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] Hereinafter, the embodiments of the determination program, the determination method, and the information processing device according to the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to these embodiments. In addition, each embodiment can be appropriately combined within a range that does not cause a contradiction.

[0010] Fig. 1 is a diagram illustrating an example of the overall configuration of a system according to Example 1. As shown in Fig. 1, this system is a system in which a server 1 and an information processing device 10 are connected to each other by wire or wirelessly.

[0011] The server 1 is an example of a computer that provides a forum for discussion in which topics are provided to a plurality of users and each user can post his / her own opinion using a social networking service, a web browser, etc. Note that the discussion is not limited to being held in real time, and for example, minutes of the discussion may be used.

[0012] The information processing device 10 is an example of a computer that analyzes opinions posted by each user and detects the majority opinion that corresponds to the majority opinion and the minority opinion that corresponds to the minority opinion, thereby generating information that will lead to follow-up with each user and stimulate future discussions.

[0013] Generally, in a discussion, there may be users who belong to the majority and users who belong to the minority. Users who belong to the minority may change their opinions due to pressure to conform from the majority and lose confidence, which causes great stress. On the other hand, it is known that users who belong to the minority and have something in common can influence and oppose the majority by banding together. In this embodiment, users who belong to the majority may be simply referred to as "majority users" and users who belong to the minority may be simply referred to as "minority users".

[0014] In other words, there are two groups of minorities in a community: "a minority group that is already united and can oppose others" and "a minority group that is not united and is unilaterally feeling stress in the community." Therefore, detecting minority users who are feeling stress and following up with them, or grouping users who are feeling similar stress to make it easier for them to participate in discussions, will lead to revitalizing discussions and providing a forum for effective discussion.

[0015] A known method for detecting minorities is negative-positive analysis, which uses the degree of negativity or positivity set for keywords in posted opinions. However, minorities rarely express their opinions except for a small number of strong minorities who can express their opinions, so negative-positive analysis makes it difficult to grasp the entirety of minority opinions. In addition, opinions that change between negative and positive are very dramatic changes, so negative-positive analysis is not suitable for observing minorities.

[0016] Therefore, the information processing device 10 according to the first embodiment formulates the behavioral characteristics of minorities based on a behavioral characteristic analysis of minorities based on social psychology, and automatically extracts minorities.

[0017] Specifically, the information processing device 10 determines the trend of change in index values ​​indicating the degree of multiple positive emotions of multiple users based on at least one of the multiple users' own opinions on a specific topic and information on the approval or disapproval of other people's opinions on the specific topic. Then, when the information processing device 10 detects a first trend and a second trend, the difference between which is equal to or less than a threshold, among the trends of change in index values ​​indicating the degree of multiple positive emotions, it outputs the first user corresponding to the first trend and the second user corresponding to the second trend among the multiple users as users included in the target group.

[0018] Fig. 2 is a diagram for explaining extraction of change in opinion according to the first embodiment. Fig. 2 shows the change in positivity by showing the transition of the most positive keyword included in an opinion posted by a certain user, or the transition of the most positive keyword included in an opinion of another person with whom a certain user agrees. For example, referring to Fig. 2, it is shown that, among the keywords included in opinions posted between time t0 and time t1, keyword A has the highest positivity (0.58) among the keywords posted around time t0.

[0019] As shown in Figure 2, from time t0 to time t4, the opinion of a certain user changes within the positive range, but from time t3 to time t4, it changes from high positivity to low positivity, although it is still positive. A typical negative-positive analysis detects dramatic changes from positive to negative, and therefore analyzes such changes within the positive as a consistent positive state.

[0020] In contrast, the information processing device 10 according to the first embodiment appropriately detects the constantly changing positive and negative reactions of minorities. In other words, the information processing device 10 detects subtle changes in the user's signal by determining the repetition of small ups and downs in the positive or negative range as the number of changes using keyword definition information. As a result, the information processing device 10 can determine multiple trends in changes in positive or negative emotions, group users who have similar changes, and identify users who have similar opinions.

[0021] 3 is a functional block diagram showing a functional configuration of the information processing device 10 according to the first embodiment. As shown in FIG.

[0022] The communication unit 11 controls communication with other devices. For example, the communication unit 11 acquires opinions (posts) of each user on a certain topic from the server 1, and transmits analysis results and the like to the administrator's terminal.

[0023] The storage unit 12 stores various data and programs executed by the control unit 20. For example, the storage unit 12 stores a user information DB13, a keyword DB14, and a posted information DB15.

[0024] The user information DB13 is a database that stores information about users who post opinions in a forum. Specifically, the user information DB13 stores information about employees who participate in a forum such as a company, and stores information about online users in the case of a SNS.

[0025] FIG. 4 is a diagram showing an example of user information stored in the user information DB13. As shown in FIG. 4, the user information DB13 stores, as an example, "user name, gender, affiliation, number of participations, field of expertise," and the like. In the "user name" stored here, the name of the user, such as real name, pen name, and name in the SNS, is registered. In the "gender," information indicating male or female is registered. In the "affiliation," the affiliation, such as company employee or department, is registered. In the "number of participations," the number of times participation in discussions is registered. In the "field of expertise," the field of expertise in which the user is good is registered. In the example of FIG. 4, the information of "user A" is registered as "male," belonging to "company A," having participated "five times," and "economics" as the field of expertise.

[0026] The keyword DB 14 is a database that stores information in which keywords are associated with index values ​​indicating positivity. Specifically, the keyword DB 14 stores a value indicating an index value calculated by a known method for each word (keyword) used in a posted opinion.

[0027] FIG. 5 is a diagram showing an example of information stored in the keyword DB 14. As shown in FIG. 5, the keyword DB 14 stores "keywords and index values" in association with each other. Here, words and parts of speech are registered as "keywords". Information that quantifies the positivity, which is the degree of a user's positive emotion, is registered as "index value". This "index value" is a value from 1 to -1, and the higher the probability that it is used for positive opinions, the higher the value. Note that the "index value" may be a value described in a dictionary for emotion analysis or a word emotion polarity correspondence table. In the example of FIG. 5, it is registered that the keyword "good" has an index value of "0.999995" and is often used for positive opinions. On the other hand, it is registered that the keyword "swear" has an index value of "-0.999961", has a high negativity, and is often used for negative opinions.

[0028] The posting information DB15 is a database that stores information about opinions posted by each user. The information stored here may be information periodically acquired by the communication unit 11, or may be information input by an administrator or the like. FIG. 6 is a diagram showing an example of information stored in the posting information DB15. As shown in FIG. 6, the posting information DB15 stores, as an example, "time, poster, posting content," etc. In the "time" stored here, the time when the opinion was posted is registered. In the "poster," information that identifies the user who posted the opinion is registered. In the "posting content," the posted content is registered. In the example of FIG. 6, it is registered that "User A" posted "Comment A" at "9:00."

[0029] The posted information DB 15 can also store information on other users who have tuned in, in association with the opinions of each user. For example, the posted information DB 15 further associates "tune-in information" in which tuned-in users X and Y are registered with "time, poster, and posted content." The control unit 20 can detect and identify tuned-in users by acquiring a button operation or the like pressed when tuning in to the opinion of a certain user.

[0030] The control unit 20 is a processing unit that controls the entire information processing device 10. For example, the control unit 20 includes a minority detection unit 21, a first detection unit 22, a second detection unit 23, and a protection determination unit 24.

[0031] The minority detection unit 21 detects minority users based on opinions posted to a topic. Specifically, the minority detection unit 21 determines whether or not each user corresponds to a minority by comparing the number of posted opinions with a threshold value, or by using the bias of the posted opinions. Then, the minority detection unit 21 detects the corresponding user as a minority user, stores it in the storage unit 12, and outputs it to another processing unit.

[0032] For example, the minority detection unit 21 detects as minority users those users whose number of posted opinions is less than a threshold, those whose posting interval is less than a threshold, those whose difference between the number of posts and the average for all users is equal to or greater than a threshold, or those whose difference between the posting interval and the average for all users is equal to or greater than a threshold.

[0033] Further, the minority detection unit 21 identifies the maximum value of the index value among the opinions posted by each user based on the information stored in the keyword DB 14. Then, the minority detection unit 21 calculates the variance and standard deviation of the maximum value of the index value of each user, and detects users whose difference between the maximum value and the variance is equal to or greater than a threshold value or whose standard deviation is equal to or greater than a threshold value as minorities. Note that instead of the maximum value, the average value of the index value of each user can also be used, and can also be combined with the number of opinions. In this way, the minority detection unit 21 can detect minority users by adopting a known method.

[0034] The first detection unit 22 observes direct changes in opinions posted by users and detects strong conformity signals corresponding to triggers or signs of conformity with other users. Specifically, the first detection unit 22 determines the tendency of changes in index values ​​indicating the degree of multiple positive emotions of multiple users based on changes in the opinions of each of the multiple users on a specific topic.

[0035] 7 is a diagram illustrating a strong synchronization signal. As shown in FIG. 7, the first detection unit 22 calculates an index of opinion change for a user who is different from the overall distribution, where the index value does not change dramatically from positive to negative, but the index value drops by a certain value within the positive range, such as when the index value changes from "+0.9" to "+0.3". Note that a user who is different from the overall distribution is an example of a minority user detected by the minority detection unit 21.

[0036] For example, the first detection unit 22 refers to the posted information DB 15 and extracts opinions posted by user X on a certain topic. The first detection unit 22 refers to the keyword DB 14 and sets the keyword α having the largest index value among the keywords included in the opinions as a value representing the opinions. This is intended to focus on opinions with strong opinions by representing a value with a large absolute value in a state where negative and positive opinions are mixed. The first detection unit 22 then sets a window (time interval T) that captures changes in opinions within a certain period T, and shifts the window by +1 along the time axis while tracing it, obtaining the degree of change in opinions before and after the time interval T and the number of changes.

[0037] FIG. 8 is a diagram illustrating the detection of a strong synchronization signal. As shown in FIG. 8, the first detection unit 22 identifies a keyword A (index value 0.58) having the maximum index value among the keywords included in the opinions posted around time t0, and identifies a keyword B (index value 0.75) having the maximum index value among the keywords included in the opinions posted around time t1. Similarly, the first detection unit 22 identifies a keyword C (index value 0.96) having the maximum index value among the keywords included in the opinions posted around time t2, and identifies a keyword D (index value 0.99) having the maximum index value among the keywords included in the opinions posted around time t3. The first detection unit 22 identifies a keyword E (index value 0.68) having the maximum index value among the keywords included in the opinions posted around time t4, and identifies a keyword F (index value 0.90) having the maximum index value among the keywords included in the opinions posted around time t5.

[0038] Then, in the range of Window "T1", the first detection unit 22 calculates the number of positive transitions to be 3 because the maximum index values ​​all rise. In the range of Window "T2" and Window "T3", the first detection unit 22 calculates the number of positive transitions to be 2 and the number of negative transitions to be 1 because the maximum index values ​​rise twice and fall once.

[0039] Then, the first detection unit 22 uses the formula 1 to calculate "S T As shown in formula (1), "S T " is the sum of the differences between the index values ​​of n opinions in the window. i+1 -α i " and "W" which indicates the sum of the number of times the index value reverses within the window. i " is calculated by multiplying it by "

[0040]

number

[0041] In addition, "W i ", as shown in formula (2), "B i " is less than 0, "-1", "B i " is 0 or more, "B i The value of "B i " is a value that changes depending on the sign of the starting point for calculating the difference and the sign of the difference, as shown in formula (3). i If the index value of " is positive (plus), then "α i+1 -α i ", starting point "α i If the index value of " is negative, use "-(α i+1 -α i )".

[0042]

number

number

[0043] Then, the first detection unit 22 calculates the "S T Then, the first detection unit 22 calculates the average value of each "S TThe first detection unit 22 calculates the variance by the formula (4) using the difference between " and the average value. The first detection unit 22 stores the calculated variance in the storage unit 12 and outputs it to the protection determination unit 24.

[0044]

number

[0045] Here, specific examples will be described with reference to Fig. 9 and Fig. 10. The first detection unit 22 can also determine the tendency of change in the index value when the index value is in a range of positive values ​​and within a range of positive opinions. As another example, the first detection unit 22 can also determine the tendency of change in the index value when the index value is in a range from positive values ​​to negative values, including a range from positive opinions to negative opinions. For this reason, specific examples of each will be described with reference to Fig. 9 and Fig. 10.

[0046] Fig. 9 is a diagram for explaining a method 1 for determining a strong tuning signal, and Fig. 10 is a diagram for explaining a method 2 for determining a strong tuning signal. Note that, although both methods are explained using an example in which there are three windows (T1, T2, T3), the number is not limited to this.

[0047] First, an example will be described with reference to Fig. 9 in which the first detection unit 22 detects a strong synchronization signal indicating a tendency of change in the index value within the range of positive opinions. The first detection unit 22 detects "S T1 ". Specifically, as shown in FIG. 9, the first detection unit 22 identifies an index value of "0.58" for the noun "safe" among keywords in opinions posted around time t0. The first detection unit 22 identifies an index value of "0.75" for the verb "understand" for opinions posted around time t1. The first detection unit 22 identifies an index value of "0.96" for the noun "correct" for opinions posted around time t2. The first detection unit 22 identifies an index value of "0.99" for the noun "appropriate" for opinions posted around time t3.

[0048] Then, the first detection unit 22 calculates the change in each window using formulas (2) and (3). Specifically, as shown in FIG. 9, the first detection unit 22 calculates the difference between the keywords "I know" and "No problem", "0.75-0.58=0.17". At this time, since the index value of the starting keyword "No problem" is "0.58" and the difference is a positive value, the first detection unit 22 sets the difference "0.17" to "W1". Similarly, the first detection unit 22 calculates the difference between the keywords "I know" and "No problem", "0.96-0.75=0.21", and since the index value of the starting keyword "I know" is "0.75" and the difference is a positive value, the first detection unit 22 sets the difference "0.21" to "W2". In addition, the first detection unit 22 calculates the difference between the keywords "appropriate" and "correct", which is "0.99-0.96=0.03", and sets the difference "0.03" to "W3" since the index value of the starting keyword "correct" is "0.96" and the difference is a positive value.

[0049] As a result, the first detection unit 22 calculates "S T1 Specifically, as shown in FIG. 9, the first detection unit 22 calculates "S T1 =(0.75-0.58)×(0.17)+(0.96-0.75)×(0.21)+(0.99-0.96)×(0.03)=0.07.

[0050] Next, the first detection unit 22 detects a change in the opinions posted by the user X for a window (T2) from time t1 to time t4 shown in FIG. T2 Specifically, as shown in FIG. 9, in addition to the above, the first detection unit 22 newly identifies an index value of “0.68” for the adverb “Tottoto” for the opinion posted around time t4.

[0051] Then, the first detection unit 22 calculates the change in each window using formula (2) and formula (3). Specifically, as shown in Fig. 9, the first detection unit 22 calculates a difference between the keywords "quickly" and "appropriate", "0.68-0.99=-0.31", in addition to the above. At this time, since the index value "0.58" of the starting keyword "appropriate" is a positive value while the difference is a negative value, the first detection unit 22 sets "-1" to "W4" based on formula (2).

[0052] As a result, the first detection unit 22 calculates "S T2 Specifically, as shown in FIG. 9, the first detection unit 22 calculates "S T2 =(0.96-0.75)×(0.21)+(0.99-0.96)×(0.03)+(0.68-0.99)×(-1)=0.36.

[0053] Next, the first detection unit 22 detects a change in the opinions posted by the user X for a window (T3) from time t2 to time t5 shown in FIG. T3 Specifically, as shown in FIG. 9, in addition to the above, the first detection unit 22 newly identifies an index value of “0.90” for the noun “solution” for the opinion posted around time t5.

[0054] Then, the first detection unit 22 calculates the change in each window using formulas (2) and (3). Specifically, as shown in Fig. 9, the first detection unit 22 calculates a difference between the keywords "solve" and "quickly", "0.90-0.68=0.22", in addition to the above. At this time, since the index value of the starting keyword "quickly" is "0.68" and the difference is a positive value, the first detection unit 22 sets the difference "0.22" to "W5".

[0055] As a result, the first detection unit 22 calculates "S T3 Specifically, as shown in FIG. 9, the first detection unit 22 calculates "S T3=(0.99-0.96)×(0.03)+(0.68-0.99)×(-1)+(0.90-0.68)×(0.22)=0.36.

[0056] After that, the first detection unit 22 detects T1 ", "S T2 ", "S T3 The first detector 22 calculates the average value of "S T1 =0.07", "S T2 =0.36", "S T3 Using "=0.36" and the mean value of "0.26", the variance "0.0187" is calculated using formula (1).

[0057] Next, an example will be described with reference to Fig. 10 in which the first detection unit 22 detects a strong synchronization signal, which is a trend of change in index values, within a range including positive and negative. The first detection unit 22 detects "S T1 ". Specifically, as shown in FIG. 10, the first detection unit 22 identifies an index value of "0.58" for the noun "safe" for opinions posted around time t0. The first detection unit 22 identifies an index value of "0.75" for the verb "understand" for opinions posted around time t1. The first detection unit 22 identifies an index value of "0.96" for the noun "correct" for opinions posted around time t2. The first detection unit 22 identifies an index value of "0.99" for the noun "appropriate" for opinions posted around time t3.

[0058] Then, the first detection unit 22 calculates the change in each window using formula (2) and formula (3). Specifically, as shown in Fig. 10, the first detection unit 22 calculates the difference "0.75-0.58=0.17" between the keywords "understand" and "safe", and sets the difference "0.17" to "W1". Similarly, the first detection unit 22 calculates the difference "0.96-0.75=0.21" between the keywords "correct argument" and "understand", and sets the difference "0.21" to "W2". In addition, the first detection unit 22 calculates the difference "0.99-0.96=0.03" between the keywords "appropriate" and "correct argument", and sets the difference "0.03" to "W3".

[0059] As a result, the first detection unit 22 calculates "S T1 Specifically, as shown in FIG. 10, the first detection unit 22 calculates "S T1 =(0.75-0.58)×(0.17)+(0.96-0.75)×(0.21)+(0.99-0.96)×(0.03)=0.07.

[0060] Next, the first detection unit 22 detects a change in the opinions posted by the user X for a window (T2) from time t1 to time t4 shown in FIG. T2 10, in addition to the above, the first detection unit 22 newly identifies an index value of "-0.99" for the adjective "noisy" for the opinion posted around time t4.

[0061] Then, the first detection unit 22 calculates the change in each window using formula (2) and formula (3). Specifically, as shown in Fig. 10, in addition to the above, the first detection unit 22 newly calculates the difference between the keywords "noisy" and "appropriate", "-0.99-0.99=1.98", and sets "-1" to "W4" based on formula (2).

[0062] As a result, the first detection unit 22 calculates "S T2 Specifically, as shown in FIG. 10, the first detection unit 22 calculates "S T2=(0.96-0.75)×(0.21)+(0.99-0.96)×(0.03)+(-0.99-0.99)×(-1)=2.02.

[0063] Next, the first detection unit 22 detects a change in the opinions posted by the user X for a window (T3) from time t2 to time t5 shown in FIG. T3 10, in addition to the above, the first detection unit 22 newly identifies an index value of "0.90" for the noun "solution" for the opinion posted around time t5.

[0064] Then, the first detection unit 22 calculates the change in each window using formula (2) and formula (3). Specifically, as shown in Fig. 10, in addition to the above, the first detection unit 22 newly calculates the difference between the keywords "solved" and "noisy", that is, "0.90-(-0.99)=1.89", and sets "-1" to "W5" based on formula (2).

[0065] As a result, the first detection unit 22 calculates "S T3 Specifically, as shown in FIG. 10, the first detection unit 22 calculates "S T3 =(0.99-0.96)×(0.03)+(-0.99-0.99)×(-1)+(0.90-(-0.99))×(-1)=0.099.

[0066] After that, the first detection unit 22 detects T1 ", "S T2 ", "S T3 The first detection unit 22 calculates the average value of "S T1 =0.07", "S T2 =2.02", "S T3 = 0.099" and the mean value of "0.73", the variance of "0.83" is calculated using formula (1).

[0067] Returning to Fig. 3, the second detection unit 23 focuses on agreement with the opinions of others, observes changes in the index values ​​of the opinions of others, and detects a weak synchronization signal corresponding to a reaction to others. Specifically, the second detection unit 23 determines the tendency of changes in the index values ​​indicating the degree of multiple positive emotions of multiple users based on information on the approval or disapproval of the opinions of others on a specific topic.

[0068] Fig. 11 is a diagram illustrating a weak conformity signal. As shown in Fig. 11, the second detection unit 23 calculates an index for an operation of conforming to another person's opinion (for example, an operation of posting a "Like" to another person's opinion) with an index value of "0.9" instead of an opinion with a negativity (synonymous with negative positivity) of "0.5" for a user who is different from the overall distribution.

[0069] For example, the second detection unit 23 refers to the posted information DB 15 and extracts opinions of others with which the user X agrees from the total number of opinions M. The second detection unit 23 refers to the keyword DB 14 and sets the keyword α having the maximum index value among the keywords included in the opinions with which the user X agrees as a value representing the opinions. In this way, the second detection unit 23 extracts the agreement flags (s j =1) or a consent flag (s j =0) is generated and stored in the storage unit 12 or the like.

[0070] Then, the second detection unit 23 uses the formula (5) to detect the consent flag (s j =1) is set to keyword β j The total value of "U j The second detection unit 23 calculates the sum "U j" is calculated. The second detection unit 23 then uses formula (6) to calculate the variance "U", which is the tendency of change in the index value indicating the degree of multiple positive emotions of multiple users. Unlike a strong conformity signal, posting of other people's opinions and agreement with other people's opinions are performed asynchronously, so no window is set. The second detection unit 23 also stores the calculated variance in the memory unit 12 and outputs it to the protection determination unit 24.

[0071]

number

number

[0072] Here, a specific example will be described with reference to Fig. 12. Fig. 12 is a diagram for explaining a method for determining a weak synchronization signal. Note that an example will be described using opinions posted up to time t5, but the time span, etc., is not limited thereto.

[0073] First, as shown in Fig. 12, the second detection unit 23 counts the total number of opinions M. The total number of opinions may be the total number of opinions of all users other than the target minority user X, or the total number including the opinions of user X, or may be the total number of opinions posted by the majority users.

[0074] Next, the second detection unit 23 identifies three opinions that the user X agrees with among the M opinions, and identifies the keyword with the maximum index value from each opinion. For example, the second detection unit 23 identifies keyword P with index value "0.60", keyword Y with index value "0.96", and keyword Z with index value "0.75".

[0075] Then, the second detection unit 23 calculates the sum of the index values ​​"0.60 + 0.96 + 0.75 = 2.31" and the average value of the index values ​​"2.31 / 3 = 0.77." Furthermore, when the total number of opinions M = 10, the second detection unit 23 calculates the variance U of "0.0654" using formula (6).

[0076] Returning to FIG. 3, the protection determination unit 24 determines whether or not there are minority users in the same circumstances as each user. Specifically, the protection determination unit 24 detects a first trend and a second trend, the difference of which is equal to or less than a threshold, among the trends of changes in index values ​​indicating the degree of a plurality of positive emotions. Then, the protection determination unit 24 outputs the first user corresponding to the first trend and the second user corresponding to the second trend as users included in the target group.

[0077] For example, the protection determination unit 24 calculates a conformity index (Conformity) for each minority user using the variance for strong conformance signals (Equation 4) and the variance for weak conformance signals (Equation 6) according to Equation (7). The protection determination unit 24 then groups users with similar conformance indices (hereinafter sometimes referred to as conforming users) into groups of users experiencing similar stress. Note that a group experiencing similar stress is an example of a group of conforming users or a group of conforming stress.

[0078]

number

[0079] FIG. 13 is a diagram for explaining detection of synchronized users. As shown in FIG. 13, the protection determination unit 24 determines that minority users whose difference in synchronization index is less than a threshold are in the same group (cohesive group). In the example of FIG. 13, the protection determination unit 24 groups users whose synchronization index is "0.7" whose difference is less than the threshold as users who feel the same stress, with respect to a determination target user whose synchronization index is "0.8". Note that the protection determination unit 24 can also determine that minority users whose overlapping number of synchronized opinions is equal to or greater than a threshold belong to the same group, and can also determine that users who satisfy both the difference in synchronization index and the overlapping number of synchronized opinions belong to the same group.

[0080] The protection determination unit 24 can also use the relationship (connection) between users as a determination criterion by referring to the user information DB 13. For example, the protection determination unit 24 can use the same gender, the same affiliation or field of expertise, or similar affiliation or field of expertise as a determination criterion.

[0081] Furthermore, the protection determination unit 24 notifies each user belonging to the same group that they are users who can post opinions in solidarity, and by making each user recognize that there are users in the same situation, it is possible to encourage more active posting of opinions than usual and to stimulate discussion. It is also expected that such users will influence the majority by uniting as a minority with something in common, and thus stimulate discussion.

[0082] In addition, the protection determination unit 24 detects a user who cannot detect other users with a similar synchronization index as an isolated user to be protected. Fig. 14 is a diagram explaining the detection of a user to be protected. As shown in Fig. 14, when the protection determination unit 24 cannot detect a user with a synchronization index whose difference is less than a threshold value for a user to be determined whose synchronization index is "0.8", the protection determination unit 24 determines the user to be determined as a protection target.

[0083] The protection determination unit 24 may also use the relationships between users as a determination material, as in Fig. 13. For example, since the protection determination unit 24 considers that the more relationships a protected user has, the stronger the sense of alienation the protected user feels, the importance of protection of the protected user may be added depending on the relationships. In this way, the protection determination unit 24 detects and lists the protected users, so that the administrator or the like can follow up on the protected users and consider measures to reduce the sense of alienation and enable even minority users to post their opinions.

[0084] Next, an example of a series of processing flow from the above-mentioned minority user detection to the detection of a protected user will be described. Fig. 15 is a flowchart showing the processing flow according to the first embodiment. Note that the order of each step in the flowchart may be changed within a range that does not cause inconsistency.

[0085] As shown in FIG. 15, when a discussion starts (S101: Yes), the control unit 20 of the information processing device 10 collects posted information including opinions posted by each user and agreement information in which each user agrees with the opinions of other users, and stores the collected information in the posted information DB 15 (S102).

[0086] Then, when the specified analysis timing is reached (S103: Yes), such as when a specified time has elapsed since the start of the discussion or when the number of posts reaches a certain number, the control unit 20 reads the posted information from the posted information DB15 and detects minority users (S104).

[0087] Here, if no minority users are detected (S105: No), the control unit 20 repeats S102 and subsequent steps. On the other hand, if minority users are detected (S105: Yes), the control unit 20 selects one of the detected minority users (S106).

[0088] Then, the control unit 20 calculates a strong synchronization signal for the selected minority user using equations (1) to (4) (S107), and calculates a weak synchronization signal using equations (5) to (6) (S108). After that, the control unit 20 calculates a synchronization index for the selected minority user using equation (7) (S109).

[0089] Here, if there is an unprocessed minority user (S110: Yes), the control unit 20 repeats S106 and subsequent steps for the next minority user. On the other hand, if there is no unprocessed minority user (S110: No), the control unit 20 uses the conformity index of each minority user to group users who share a common conformity stress (S111), and identifies isolated users to be protected (S112).

[0090] As described above, the information processing device 10 can generate multiple trends, such as strong and weak matching signals, regarding changes in index values ​​indicating the degree of positive emotions of users of each minority. Furthermore, the information processing device 10 can group users who have the same tendency by comparing the multiple trends. As a result, the information processing device 10 can identify users who have similar opinions.

[0091] In addition, the information processing device 10 can calculate a strong conformity signal obtained by directly analyzing the user's opinions and a weak conformity signal obtained by analyzing the user's reactions to the opinions of others. As a result, the information processing device 10 can perform grouping that takes into account both positive and negative aspects of each user.

[0092] The information processing device 10 can also generate and output a list of strong synchrony signals calculated for each user, a list of weak synchrony signals calculated for each user, a list of groupings, a list of users to be protected, etc. As a result, the information processing device 10 can output useful information that can be used by the administrator in taking measures to stimulate discussions or to protect minorities.

[0093] The data examples, numerical values, keywords, information of each DB, etc. used in the above embodiment are merely examples and can be changed as desired. In addition, the above embodiment has been described using an example of posting opinions using SNS, etc., but is not limited to this. For example, the information processing device 10 can perform morphological analysis or document analysis on minutes of a documented discussion or audio data of recorded voice, and identify the opinions of each user, thereby performing an analysis similar to the above process. In this case, the information processing device 10 can also use, for example, raising a hand, a backchannel, or the time it takes to reply to an opinion as a pro or con of another person's opinion.

[0094] In the above embodiment, the information processing device 10 identifies minority users and then performs analysis on the minority users, but the present invention is not limited to this. For example, the information processing device 10 can also perform analysis on all users who have posted.

[0095] In addition, the information processing device 10 can calculate each synchronization signal and each synchronization index using only the change in the positivity within the positives, and can calculate each synchronization signal and each synchronization index using only the change in the negativity within the negatives.

[0096] Furthermore, the information processing device 10 can group users and identify users to be protected by using either a strong or weak assimilation signal. In the above embodiment, the information processing device 10 uses the positivity of the opinions of others with whom the target user agrees to calculate the weak assimilation signal, but the present invention is not limited to this. For example, the information processing device 10 can also use the positivity of opinions of others other than those with which the target user does not agree. Furthermore, when agreement is divided into multiple levels, the information processing device 10 can also use the positivity of opinions of others with which a predetermined level of agreement or higher has been achieved.

[0097] In the above embodiment, the higher the numerical value, the higher the positivity, but the present invention is not limited to this, and the information processing device 10 can adopt any information that can express positivity, regardless of the format. For example, even if the higher the numerical value, the higher the negativity, the information processing device 10 can execute an analysis using the same processing as the above embodiment by using absolute values ​​or code conversion. As a code conversion method, for example, when the index value is a negativity of "-0.99", the information processing device 10 can adopt an analysis in which the negative negativity (0.99) is the positivity.

[0098] The information processing device 10 can also perform analysis using negativity instead of positivity. Even if negativity is set as an index value, the information processing device 10 can perform analysis similar to that of the above embodiment by processing negative negativity as positivity.

[0099] The information including the processing procedures, control procedures, specific names, various data and parameters shown in the above documents and drawings may be changed arbitrarily unless otherwise specified.

[0100] Furthermore, the specific form of distribution or integration of the components of each device is not limited to that shown in the figure. For example, the first detection unit 22 and the second detection unit 23 may be integrated. In other words, all or part of the components may be functionally or physically distributed or integrated in any unit depending on various loads and usage conditions. Furthermore, all or any part of the processing functions of each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware using wired logic.

[0101] Fig. 16 is a diagram for explaining an example of a hardware configuration. As shown in Fig. 16, an information processing device 10 includes a communication device 10a, a hard disk drive (HDD) 10b, a memory 10c, and a processor 10d. The components shown in Fig. 16 are connected to each other via a bus or the like.

[0102] The communication device 10a is a network interface card or the like, and communicates with other devices. The HDD 10b stores programs and DBs that operate the functions shown in FIG.

[0103] The processor 10d reads out a program that executes the same processes as the respective processing units shown in FIG. 3 from the HDD 10b, etc., and loads it in the memory 10c, thereby operating a process that executes each function described in FIG. 3, etc. For example, this process executes a function similar to that of each processing unit possessed by the information processing device 10. Specifically, the processor 10d reads out a program having a function similar to that of the minority detection unit 21, the first detection unit 22, the second detection unit 23, the protection determination unit 24, etc., from the HDD 10b, etc. Then, the processor 10d executes a process that executes a process similar to that of the minority detection unit 21, the first detection unit 22, the second detection unit 23, the protection determination unit 24, etc.

[0104] In this way, the information processing device 10 operates as an information processing device that executes a determination method by reading and executing a program. The information processing device 10 can also realize functions similar to those of the above-mentioned embodiment by reading the program from a recording medium using a medium reading device and executing the read program. Note that the program in this other embodiment is not limited to being executed by the information processing device 10. For example, the above-mentioned embodiment may be similarly applied to a case where another computer or server executes a program, or a case where these execute a program in cooperation with each other.

[0105] This program may be distributed via a network such as the Internet. Also, this program may be recorded on a computer-readable recording medium such as a hard disk, a flexible disk (FD), a CD-ROM, a magneto-optical disk (MO), or a digital versatile disk (DVD), and may be read out from the recording medium and executed by a computer. [Explanation of symbols]

[0106] 10. Information processing device 11 Communications Department 12 Storage section 13 User information DB 14 Keyword Database 15 Posting Information DB 20 Control section 21 Minority Detection Unit 22 First detection unit 23 Second detection section 24 Protection judgment section

Claims

1. When using information regarding the opinions of each of a plurality of users regarding a specific topic, index values ​​indicating the degree of a plurality of positive emotions of the plurality of users are identified based on keywords contained in the opinions of the users. When using information regarding the approval or disapproval of the opinions of others regarding the specific topic, the index values ​​are identified based on keywords contained in the opinions of the others with which the users agree, and a trend in changes in the index values ​​is determined based on the number of changes in the index values ​​at regular intervals or within a regular interval. when a first trend and a second trend, the difference between which is equal to or less than a threshold, are detected from among the trends of changes in index values ​​indicating the degrees of the plurality of positive emotions, a first user corresponding to the first trend and a second user corresponding to the second trend among the plurality of users are output as users included in a target group; A determination program that causes a computer to execute a process.

2. Calculating a variance using the index value specified based on the number of times the person has posted an opinion on the specific topic, the interval at which the person has posted an opinion on the specific topic, or a keyword included in the person's opinion on the specific topic; extracting, from among target users who have opinions on the specific topic based on the calculated information, users whose opinions are in the minority on the specific topic as the plurality of users; 2. The determination program according to claim 1, which causes the computer to execute a process.

3. For each of the plurality of tendencies determined for each of the plurality of users, it is determined whether or not the tendency of another user in which the difference is equal to or smaller than a threshold has been detected; extracting users from the plurality of users for whom the tendency of the other users to have the difference equal to or smaller than the threshold is not detected; 2. The determination program according to claim 1, which causes the computer to execute a process.

4. The process of determining includes: When using information about the user's opinion on the specific topic, for each of the users, the index value based on a keyword included in the user's opinion and the number of times the index value has changed within a range indicating a positive opinion are identified at regular intervals; calculating a variance of the index value as a trend of change of the index value by using the index value and the number of changes at each fixed time; The output process includes: using each variance calculated for each of the plurality of users, outputting the first user and the second user whose difference in variance is less than a threshold as users included in the target group; 2. The determination program according to claim 1.

5. The process of determining includes: When using information about the user's opinion on the specific topic, for each of the users, identify the index value based on a keyword included in the user's opinion and the number of times the index value has changed within a certain period of time from a negative opinion to a positive opinion; calculating a variance of the index value as a trend of change of the index value by using the index value and the number of changes at each fixed time; The output process includes: using each variance calculated for each of the plurality of users, outputting the first user and the second user whose difference in variance is less than a threshold as users included in the target group; 2. The determination program according to claim 1.

6. The process of determining includes: In the case where information on the approval or disapproval of the opinions of others regarding the specific topic is used, for each of the plurality of users, a total number of opinions of the plurality of users and the index value based on keywords included in the opinions of others with which the plurality of users agree are identified within a certain period of time; calculating a variance of the index value as a trend of change of the index value by using the total number of opinions and the index value; The output process includes: using each variance calculated for each of the plurality of users, outputting the first user and the second user whose difference in variance is less than a threshold as users included in the target group; 2. The determination program according to claim 1.

7. When using information regarding the opinions of each of a plurality of users regarding a specific topic, index values ​​indicating the degree of a plurality of positive emotions of the plurality of users are identified based on keywords contained in the opinions of the users. When using information regarding the approval or disapproval of the opinions of others regarding the specific topic, the index values ​​are identified based on keywords contained in the opinions of the others with which the users agree, and a trend in changes in the index values ​​is determined based on the number of changes in the index values ​​at regular intervals or within a regular interval. when a first trend and a second trend, the difference between which is equal to or less than a threshold, are detected from among the trends of changes in index values ​​indicating the degrees of the plurality of positive emotions, a first user corresponding to the first trend and a second user corresponding to the second trend among the plurality of users are output as users included in a target group; A determination method characterized in that the processing is executed by a computer.

8. When using information regarding the opinions of each of a plurality of users regarding a specific topic, index values ​​indicating the degree of a plurality of positive emotions of the plurality of users are identified based on keywords contained in the opinions of the users. When using information regarding the approval or disapproval of the opinions of others regarding the specific topic, the index values ​​are identified based on keywords contained in the opinions of the others with which the users agree, and a trend in changes in the index values ​​is determined based on the number of changes in the index values ​​at regular intervals or within a regular interval. when a first trend and a second trend, the difference between which is equal to or less than a threshold, are detected from among the trends of changes in index values ​​indicating the degrees of the plurality of positive emotions, a first user corresponding to the first trend and a second user corresponding to the second trend among the plurality of users are output as users included in a target group; An information processing device comprising a control unit.

Images