Patent Application
20130117296
The present invention relates to a communication assistance device, a communication assistance method, and a computer-readable recording medium having recorded therein a program for realizing the communication assistance device and the communication assistance method.
In recent years, communication assistance systems have been proposed. Communication assistance systems assist development of human relationships based on how close hobbies and preferences of one person to those of another. For example, Non-Patent Document 1 discloses one example of conventional communication assistance systems.
The communication assistance system disclosed in Non-Patent Document 1 aims to make two people acquainted with each other over time. More specifically, this communication assistance system first searches for two people with similar preferences and an agent who can mediate these two people. Here, whether or not their preferences are similar is determined based on their life patterns in the past, e.g. whether or not they have visited the same place.
This communication assistance system then judges whether or not the three people who have been searched for are at the same place, and when judging that the three people are at the same place, urges the agent to mediate the two people with similar preferences.
In this way, the communication assistance system disclosed in Non-Patent Document 1 provides an opportunity for the two people with similar preferences to be introduced to each other via the agent. This makes it easy for the two people who are introduced to each other to communicate with each other, compared to the case where they have to initiate conversation with a stranger who shares similar interests.
Although the communication assistance system disclosed in Non-Patent Document 1 aims to assist communication in a real space, it can likewise assist communication in a virtual space such as a social networking service (SNS).
Non-Patent Document 1:
However, the communication assistance system disclosed in Non-Patent Document 1 has the problem that it does not grasp human relationships in a stepwise manner and therefore cannot provide communication assistance in accordance with the levels of human relationships.
More specifically, the communication system disclosed in Non-Patent Document 1 attempts to give two people a chance to communicate with each other via an agent. At this time, the two people are suddenly introduced to and encouraged to befriend each other although they are strangers, i.e. they do not know each other at all. The two people may thus end up being unfamiliar with each other, with the result that their communication does not last long. The same goes for communication in a virtual space.
Level-based communication assistance can be provided as long as human relationships, such as strangers, strangers with similar preferences, people with similar preferences and interest in each other, and friends, can be grasped in a stepwise manner. That is to say, communication assistance is required that takes into consideration the intermediate levels between the level at which the two people are strangers and the level at which they become friends.
With consideration of such intermediate levels, the two people can get to know each other and cultivate a human relationship. Furthermore, in this case, even if users move through different levels and end up having similar preferences, they will not suddenly jump to the level at which they become friends, and accordingly a human relationship can be reliably cultivated. In view of the above, in a communication assistance system, it is important to determine communicational human relationships in a stepwise manner and provide level-based assistance.
One example of the aims of the present invention is to provide a communication assistance device, a communication assistance method and a computer-readable recording medium that can solve the above problem by determining a level of a relationship between users who communicate with each other.
In order to achieve the above aim, a communication assistance device according to one aspect of the present invention includes: a communication level determination unit that determines a level of a relationship between users based on similarity between the users and on user action records, the similarity between the users being obtained from preference information showing preferences of the users, and the user action records showing records of actions taken by a certain user toward a partner user with whom the certain user communicates out of the users.
Furthermore, in order to achieve the above aim, a communication assistance method according to one aspect of the present invention includes: a communication level determination step of determining a level of a relationship between users based on similarity between the users and on user action records, the similarity between the users being obtained from preference information showing preferences of the users, and the user action records showing records of actions taken by a certain user toward a partner user with whom the certain user communicates out of the users.
Furthermore, in order to achieve the above aim, a computer-readable recording medium according to one aspect of the present invention has recorded therein a program including instructions for causing a computer to execute: a communication level determination step of determining a level of a relationship between users based on similarity between the users and on user action records, the similarity between the users being obtained from preference information showing preferences of the users, and the user action records showing records of actions taken by a certain user toward a partner user with whom the certain user communicates out of the users.
As set forth above, a communication assistance device, a communication assistance method and a computer-readable recording medium according to the present invention can determine a level of a relationship between users who communicate with each other.
The following describes a communication assistance device, a communication assistance method and a program according to Embodiment 1 of the present invention with reference to
As shown in
As shown in
In this manner, the communication assistance device 10 can determine a level of a relationship between users who communicate with each other. The communication assistance device 10 can provide communication assistance to users who communicate eath other via the service providing device 30, in accordance with the determined level. By thus providing level-based communication assistance, the relationship between the users can be cultivated over time.
The following is a more specific description of the configuration of the communication assistance device 10. As shown in
The input reception unit 12 receives, as the input, a user name from the service providing device 30. In the present Embodiment 1, the service providing device 30 inputs at least one user name of a user for whom a communication level is to be determined (the aforementioned certain user). When the user name is input, the communication level determination unit 11 determines a communication level between the user corresponding to the input user name (hereinafter referred to as “input user”) and another user with whom the input user communicates (hereinafter referred to as “partner user”), and accumulates the determined communication level in the communication level accumulation unit 13.
The preference similarity accumulation unit 14 accumulates preference similarities between users as will be described later. User models, which will be described later with reference to
More specifically, the communication level determination unit 11 first refers to the user action records accumulated in the user action record accumulation unit 15 and obtains the types and frequencies of actions that the input user has taken toward a partner user. Based on the types and frequencies of the actions, the communication level determination unit 11 calculates a user interest degree which is a degree of interest that the input user takes in the partner user. Furthermore, the communication level determination unit 11 refers to the preference similarities between users accumulated in the preference similarity accumulation unit 14 and obtains preference similarity indicating a degree of similarity between preferences of the input user and preferences of the partner user.
Thereafter, the communication level determination unit 11 refers to communication levels that have been accumulated in the communication level accumulation unit 13 up to the current time. Based on the obtained preference similarity and the calculated user interest degree, the communication level determination unit 11 judges whether the relationship between the input user and the partner user is in an advancing state, a stalled state, or a receding state. The communication level determination unit 11 then updates the communication level using the result of the judgment and accumulates the updated communication level in the communication level accumulation unit 13.
Note that communication levels represent relationships between users, and users move through different communication levels in a stepwise manner. Examples of communication levels include a stranger level, an intermediate level between strangers and friends, and a friend level. The intermediate level may be classified into a plurality of levels. In the present case, the following levels (A1) to (A4) may be used as communication levels.
(A1) Stranger level: the input user is not similar to the partner user in preferences and is not interested in the partner user.
(A2) Stranger level with similar preferences and no knowledge of name: the input user is similar to the partner user in preferences but is not interested in the partner user.
(A3): Candidate-for-friend level with similar preferences: the input user is similar to the partner user in preferences and is interested in the partner user.
(A4) Friend level: the input user and the partner user have similar preferences and actively exchange information with each other.
Note that the aforementioned term “no knowledge of name” denotes the state where the input user is aware of the existence of the partner user but does not know the user name of the partner user.
Classification of communication levels may be even more specific. For example, the above level (A2), i.e. the stranger level with similar preferences, and the above level (A3), i.e. the candidate-for-friend level with similar preferences, may each be classified into an existence awareness level at which the input user is aware of the existence of the partner user and an information exchange level at which the input user exchanges information with the partner user. In this case, the following levels (B1) to (B6) may be used as communication levels.
(B1) Stranger level: the input user is not similar to the partner user in preferences and is not interested in the partner user.
(B2) Existence awareness level at stranger level with similar preferences and no knowledge of name: the input user is similar to the partner user in preferences, is not interested in the partner user, and is aware of the existence of the partner user.
(B3) Information exchange level at stranger level with similar preferences and no knowledge of name: the input user is similar to the partner user in preferences, is not interested in the partner user, and exchanges information with the partner user.
(B4): Existence awareness level at candidate-for-friend level with similar preferences: the input user is similar to the partner user in preferences, is interested in the partner user, and is aware of the existence of the partner user.
(B5): Information exchange level at candidate-for-friend level with similar preferences: the input user is similar to the partner user in preferences, is interested in the partner user, and exchanges information with the partner user.
(B6) Friend level: the input user and the partner user have similar preferences and actively exchange information with each other.
In the present Embodiment 1, the preference similarity accumulation unit 14 accumulates preference similarities between all users of the service provided by the service providing device 30. The preference similarities are calculated based on similarity between preference information of one user and preference information of another user. In the present embodiment, upon obtaining preference information of users, the service providing device 30 outputs the obtained preference information to the preference similarity accumulation unit 14. Then, based on the output preference information of users, the preference similarity accumulation unit 14 calculates preference similarities between the users and accumulates the calculated preference similarities.
Note that preference information of a user denotes information showing what the user is interested in and curious about. Examples of the preference information include a group of terms showing what the user is interested in, a group of news articles about which the user is curious, and places where the user frequently visits.
The more similar preference information of one user is to preference information of another user, the larger value the calculated preference similarity between the users has. For example, preference similarity between users can be obtained by generating term vectors based on groups of terms in which the users are interested and calculating cosine similarity between the term vectors. In this case, the cosine similarity between the term vectors is used as the preference similarity between the users. Alternatively, preference similarity between users may be obtained by generating news article vectors based on groups of news articles about which the users are curious and calculating cosine similarity between the news article vectors. In this case, the cosine similarity between the news article vectors is used as the preference similarity between the users.
Furthermore, in the present Embodiment 1, the user action record accumulation unit 15 accumulates records of actions that a certain user has taken toward another user, e.g. identifiers for identifying the actions and times when the actions were taken. Note that in the present Embodiment 1, actions denote operations performed by users via the terminal devices 20. In particular, actions include operations performed by a certain user with respect to another user, and operations equivalent to responses with respect to operations performed by another user.
More specifically, operations performed by a certain user with respect to another user include viewing of profile information, viewing of diaries, comments to diaries, transmission of messages, disclosure of a name, and the like. On the other hand, operations equivalent to responses with respect to operations performed by another user include viewing of transmitted messages, transmission of return messages, and the like.
A description is now given of the operations of the communication assistance device 10 according to Embodiment 1 of the present invention with reference to
First, as shown in
For example, in the case where the service providing device 30 inputs a user name of user A, the input reception unit 12 considers user A as an input user. Note that this is merely an example. Alternatively, in order to determine a communication level between a certain user and his/her partner user, the input reception unit 12 may receive a user name of the certain user and a user name of the partner user in a pair.
Next, the communication level determination unit 11 refers to the user action records accumulated in the user action record accumulation unit 15 and obtains user action records corresponding to the user name of the input user (step S2).
The following describes examples of the user action records accumulated in the user action record accumulation unit 15 with reference to
As shown in
For example, in
Thereafter, the communication level determination unit 11 calculates user interest degrees which are degrees of an interest that the input user takes in partner users (step S3). In step S3, the communication level determination unit 11 calculates the user interest degrees based on the types of the actions that the input user took toward the partner users, the number of times the actions were taken, and the times when the actions were taken. The more the input user is interested in the partner users, the larger value the calculated user interest degrees have.
For example, provided that a user interest degree of the input user with respect to a partner user is expressed as “C (input user, partner user)”, the communication level determination unit 11 can calculate a user interest degree using the following Expression 1.
C(input user,partner user)=AT(input user,partner user)/ATv (Expression 1)
In the above Expression 1, “AT (input user, partner user)” denotes the number of times weighted actions were taken in a predetermined time period, that is, the number of times the input user took weighted actions toward a partner user in a predetermined time period. Note that “AT (input user, partner user)” can be calculated using the following Expression 2.
Also, in the above Expression 1, “ATv” denotes the average of the numbers of times the weighted actions were taken toward all users in the predetermined time period. In other words, “ATv” can be obtained by calculating the numbers of times the weighted actions were taken toward all users and calculating the average thereof.
AT(input user,partner user)=Σ(α(action type)×T(input user,partner user,action type)) (Expression 2)
In the above Expression 2, “α (action type)” denotes a weight for each action type. Also, “T (input user, partner user, action type)” denotes the number of times the input user took actions of that action type toward a partner user in the predetermined time period.
Therefore, the user interest degree expressed by the above Expression 1 is equivalent to a ratio between the total number of times the input user took weighted actions toward a partner user in a predetermined time period and the average of the numbers of times actions were taken toward all partner users in the predetermined time period.
Also, “α (action type)” in the above Expression 2 is set for each action type. The higher the level of an action as a means of communication between users, the larger the weight for that action. For example, comparing the viewing of profile information with message transmission, the viewing of profile information is nothing but a simple act of viewing, while message transmission requires contacting a partner user. Therefore, a larger weight is given to the former. More specifically, three is set as the weight for message transmission, and one is set as the weight for profile viewing.
In the case of the examples of
In the above Expressions 1 and 2, the “predetermined time period” denotes a time period that has been set in advance for determining a communication level. The “predetermined time period” may be, for example, two days. Furthermore, the “predetermined time period” may be fixed in the communication assistance device 10, or may be a time period from when a communication level was previously determined to the current time.
Note that the user interest degree calculated using the above Expressions 1 and 2 in the present Embodiment 1 is one example. In the present Embodiment 1, a user interest degree may be calculated in consideration of ease in communication with users.
For example, a user interest degree may be calculated by dividing the number of times the input user took actions toward a partner user in a predetermined time period by the number times the input user took actions toward all partner users over an entire time period. Alternatively, the user interest degree calculated using any of the aforementioned methods may be divided by the number of times the input user took weighted actions toward all partner users over an entire time period, and the value obtained through this division may be used as the user interest degree.
In the case where the user interest degrees are calculated in the above manner, if more actions are taken toward a user with whom communication can be easily conducted than other users, the user interest degree calculated with respect to that user has a large value. On the other hand, the user interest degree calculated with respect to a user with whom communication cannot be easily conducted has a large value even if fewer actions are taken toward that user.
Furthermore, a time-based weight may be applied to calculation of the user interest degrees. For example, when a long predetermined time period is set, a weight may be adjusted depending on the time when each action was taken. More specifically, a small weight may be applied to old actions, and a large weight may be applied to recent actions. Furthermore, a weight for each action may be changed in accordance with a current communication level.
The following describes the case where user interest degrees C. are calculated for the examples of
C(user A,user B)={α(diary viewing)×T(user A,user B,diary viewing)+α(message transmission)×T(user A,user B,message transmission)}/ATv (Expression 3)
Provided that α (profile viewing)=1, α (diary viewing)=2, and α (message transmission)=3, the above Expression 3 is substituted with the following Expression 4.
C(user A,user B)=(2×1+3×1)/ATv (Expression 4)
ATv (the average of the numbers of times weighted actions were taken toward all users in a predetermined time period) is calculated as follows. It is assumed that users A, B and C compose all users. The following Expressions 5 through 10 show AT (input user, partner user) for the cases where different users act as the input users.
AT(user A,user B)=(2×1+3×1)=5 (Expression 5)
AT(user A,user C)=1×1=1 (Expression 6)
AT(user B,user A)=0 (Expression 7)
AT(user B,user C)=0 (Expression 8)
AT(user C,user A)=1×1=1 (Expression 9)
AT(user C,user B)=0 (Expression 10)
Based on the values obtained from the above Expressions 5 through 10, ATv can be calculated as shown in the following Expression 11, and C (user A, user B) can be calculated as shown in the following Expression 12.
ATv=(5+1+0+0+1+0)/6=7/6=approximately 1.0 (Expression 11)
C(user A,user B)=(2×1+3×1)/1.0=5.0 (Expression 12)
Similarly, provided that users A and C are the input user and the partner user, respectively, a user interest degree C. (user A, user C) can be calculated using the following Expression 13.
C(user A,user C)={α(profile viewing)×T(user A,user B,profile viewing)}/ATv=(1×1)/1.0=1.0 (Expression 13)
Furthermore, provided that users B and A are the input user and the partner user, respectively, similar calculation leads to a user interest degree C. (user B, user A)=0. Furthermore, provided that users C and A are the input user and the partner user, respectively, similar calculation leads to a user interest degree C. (user C, user A)=1.0.
It is apparent from the above examples that a user interest degree tends to be higher when a higher-level communication means such as message transmission has been implemented, and also when actions have been taken a larger number of times.
Next, the communication level determination unit 11 refers to preference similarities between users accumulated in the preference similarity accumulation unit 14, and obtains preference similarities between the input user and all other users (step S4).
The following describes examples of the preference similarities accumulated in the preference similarity accumulation unit 14 with reference to
In the examples of
For example, provided that user A is the input user, the preference similarity accumulation unit 14 obtains the first record and the second record (the first row and the second row). Note that in the present Embodiment 1, the obtainment of the preference similarities in step S4 may be performed in parallel with or prior to the calculation of the user interest degrees in step S3.
Thereafter, the communication level determination unit 11 determines a communication level based on the user interest degrees calculated in step S3 and the preference similarities obtained in step S4 (step S5). In the present Embodiment 1, the communication level determination unit 11 determines a communication level by judging whether the current communication level is in an advancing state, a stalled state, or a receding state.
More specifically, the communication level determination unit 11 refers to the communication levels accumulated in the communication level accumulation unit 13 and obtains a current communication level.
Examples of the communication levels accumulated in the communication level accumulation unit 13 are described below with reference to
As stated earlier, communication levels represent relationships between users, and users move through different communication levels in a stepwise manner. Examples of the communication levels include (A1) to (A4) described above.
In step S5, the communication level determination unit 11 can judge whether the communication level is in the advancing state, the stalled state or the receding state in the following manner using the user interest degrees and the preference similarities between users.
For example, the communication level determination unit 11 judges that the communication level is in the advancing state when the preference similarity is equal to or greater than a first predetermined value set therefor and the user interest degree is equal to or greater than a first predetermined value set therefor. On the other hand, the communication level determination unit 11 judges that the communication level is in the receding state when the preference similarity between users is equal to or smaller than a second predetermined value set therefor or when the user interest degree is equal to or smaller than a second predetermined value set therefor. Furthermore, the communication level determination unit 11 judges that the communication level is in the stalled state when the communication level is neither in the advancing state nor in the receding state.
A communication level between user A and user B is discussed below. In the pre-update state shown in
It is assumed that a communication level is in the advancing state when the preference similarity between users is equal to or greater than 0.5 and the user interest degrees of both users are equal to or greater than 1. It is also assumed that a communication level is in the receding state when the preference similarity between users is equal to or smaller than 0.2 and the user interest degree of one of the users is equal to or smaller than 1. It is also assumed that a communication level is in the stalled state in cases other than the above. In view of the above, in the post-update state, a communication level between user A and user B is judged to be in the stalled state. That is to say, the communication level between user A and user B does not progress from the current level, namely the candidate-for-friend level with similar preferences.
A communication level between user A and user C is discussed below. In the pre-update state shown in
As set forth above, in the present Embodiment 1, even if there is a close similarity between users, the users are first judged to be at an intermediate level instead of suddenly being judged to be at a friend level. Therefore, the relationship between the users can be cultivated.
The judgment as to whether a communication level is in the advancing state, the stalled state or the receding state may vary depending on communication levels. For example, assume that the communication levels are more specifically classified into (B1) to (B6) described above. In this case, provided that a current communication level is a stranger level, the judgment may be made as follows.
When the preference similarity between users is equal to or greater than a first predetermined value set therefor and a user interest degree of one of the users is equal to or greater than a first predetermined value set therefor, the communication level is judged to be in the advancing state. When the preference similarity between users is equal to or smaller than a second predetermined value set therefor or when a user interest degree of one of the users is equal to or smaller than a second predetermined value set therefor, the communication level is judged to be in the receding state. In cases other than the above, the communication level is judged to be in the stalled state.
When the communication levels are specifically classified, it is preferable that users move to the next level (the existence awareness level at the stranger level) if one user is interested in the other in the above manner.
On the other hand, when the current communication level is the existence awareness level at the stranger level, it is preferable to judge that the communication level is in the advancing state when the preference similarity between users is equal to or greater than a predetermined value set therefor and the user interest degrees of both users are equal to or greater than a predetermined value set therefor. In this way, when both users are interested in each other and exchange information by any means, they can move to the next level, namely the information exchange level at the stranger level.
Once step S5 has been executed in the above manner, the communication level determination unit 11 updates the determined communication level as well as the corresponding preference similarity between users and the corresponding user interest degree in the communication level accumulation unit 13 as shown in
As set forth above, the operations of the communication assistance device 10 allow determining a level of a relationship between users based on similarities between preference patterns of the users and records of actions taken by a user to other users. The communication assistance device 10 can provide communication assistance by determining the level in accordance with the level.
When users are at a stranger level with similar preferences and no knowledge of name, the communication assistance device 10 can instruct the service providing device 30 not to disclose the names of the users. If the communication level progresses to a candidate-for-friend level, the communication assistance device 10 allows the service providing device 30 to disclose the names of the users. Through the above processing, users can communication with one another while overcoming psychological obstacles over time.
It suffices for a program according to the present Embodiment 1 to cause a computer to execute steps S1 to S5 shown in
The preference similarity accumulation unit 14 is realized by the CPU and a storage device, such as a hard disk, provided in the computer. Furthermore, the communication level accumulation unit 13 and the user action record accumulation unit 15 are realized by the storage device.
The following describes a communication assistance device, a communication assistance method and a program according to Embodiment 2 of the present invention with reference to
As shown in
In the present Embodiment 2, the service providing device 30 has a function of recommending actions to users unlike Embodiment 1. The users can take the recommended actions where necessary via the terminal devices 20.
The communication level determination unit 51 receives a user name as the input, determines a communication level between an input user and a partner user, and accumulates the determined communication level in the communication level accumulation unit 13. At this time, the communication level determination unit 51 accesses the recommended action record accumulation unit 52.
More specifically, the communication level determination unit 51 first refers to the user action records accumulated in the user action record accumulation unit 15 and the recommended action records accumulated in the recommended action record accumulation unit 52. Next, the communication level determination unit 51 obtains the types and frequencies of actions that the input user took toward partner users, as well as the types and frequencies of actions that were recommended by the service providing device 30 to users. Based on the obtained types and frequencies of actions, the communication level determination unit 51 calculates user interest degrees which are degrees of an interest that the input user takes in the partner users.
In the present Embodiment 2, the recommended action record accumulation unit 52 accumulates records of actions that the service providing device 30 recommended to users. Examples of the actions that the service providing device 30 recommends to users include actions that are expected to be taken by the users to assist communication.
The following describes an example of a method used by the service providing device 30 to recommend actions to users. For example, when a user calls a certain operation, the service providing device 30 receives the called operation as the input, identifies the next action corresponding to the input, and recommends the identified action.
More specifically, when user A calls an operation for viewing a diary of user B, the service providing device 30 recommends, to user A, an action for writing a comment to the diary of user B. When user A calls an operation for storing information about which user B is significantly curious, the service providing device 30 recommends, to user A, an action for transmitting that information to user B.
As shown in
A description is now given of the operations of the communication assistance device 50 according to Embodiment 2 of the present invention with reference to
As shown in
First, steps S11 and S12 are executed. Steps S11 and S12 are similar to steps S1 and S2 in
Next, the communication level determination unit 51 refers to the recommended action records accumulated in the recommended action record accumulation unit 52, and obtains recommended action records corresponding to the user name input in step S11 (step S13).
For example, provided that user A is the input user, the communication level determination unit 51 obtains recommended action records corresponding to user A, namely recommended actions of the second, third, fourth and fifth rows (see
Thereafter, the communication level determination unit 51 calculates user interest degrees of the input user toward partner users (step S14). In the present Embodiment 2, the communication level determination unit 51 calculates the user interest degrees in step S14 using the recommended action records in addition to the types of actions that the input user took toward the partner users, the number of times the actions were taken, and the times when the actions were taken. In the present Embodiment 2 also, the more the input user is interested in partner users, the larger value the calculated user interest degrees have.
For example, in the present Embodiment 2, provided that a user interest degree of the input user with respect to a partner user is expressed as “C (input user, partner user)”, the communication level determination unit 51 can calculate the user interest degree using the following Expression 14.
C(input user,partner user)=AT(input user,partner user)/ATv+RT(input user,partner user)/RTv (Expression 14)
In the above Expression 14, “AT (input user, partner user)” denotes the number of times the input user took weighted actions toward a partner user in a predetermined time period as with Expression 1. Note that “AT (input user, partner user)” can be calculated using the above Expression 2. Furthermore, “ATv” denotes the average of the numbers of times weighted actions were taken toward all users in a predetermined time period as with Expression 1.
On the other hand, in the above Expression 14, “RT (input user, partner user)” denotes the number of times weighted recommended actions were taken in a predetermined time period, namely the number of times the input user took weighted recommended actions toward a partner user in a predetermined time period. Note that “RT (input user, partner user)” can be calculated using the following Expression 15. Furthermore, “RTv” denotes the average of the numbers of times the weighted recommended actions were taken toward all users in a predetermined time period, that is, the average of the results of calculating the numbers of times the weighted recommended actions were taken toward all users.
RT(input user,partner user)=Σ(α(action type)×Tr(input user,partner user,action type)) (Expression 15)
In the above Expression 15, “α (action type)” denotes a weight for each action type as with “α (action type)” in the above Expression 2. Furthermore, “Tr (input user, partner user, action type)”, which can be calculated by comparing the recommended action records with the user action records, denotes the number of times the input user took recommended actions of that action type in a predetermined time period.
Therefore, as can be understood from the above Expressions 14 and 15, a user interest degree according to the present Embodiment 2 can be obtained by adding a ratio between the number of times the input user took actions toward a partner user in a predetermined time period and the corresponding average, and a ratio between the number of times the input user actually took recommended actions out of the actions recommended by the service providing device 30 and the corresponding average. Therefore, a user interest degree serves as the index showing to what extent actions were taken toward an average user and to what extent a user took recommended actions.
For example, assume the case where user interest degrees are calculated based on the recommended action records shown in
RT(user A,user B,message transmission)=1 (Expression 16)
RT(user A,user C,profile viewing)=1 (Expression 17)
RT(user C,user A,profile viewing)=1 (Expression 18)
Based on the examples of
C(user A,user B)={(α(diary viewing)×T(user A,user B,diary viewing)+α(message transmission)×T(user A,user B,message transmission))}/ATv+{α(message transmission)×Tr(user A,user B,message transmission)}/RTv (Expression 19)
Provided that α (profile viewing)=1, α (diary viewing)=2, and α (message transmission)=3, the above Expression 19 is substituted with the following Expression 20 based on the results of the above Expressions 16, 3 and 4.
C(user A,user B)=(2×1+3×1)/ATv+(3×1)/RTv (Expression 20)
RTv (the average of the numbers of times the weighted recommended actions were taken toward all users in a predetermined time period) is calculated as follows. It is assumed that users A, B and C compose all users. The following Expressions 21 through 26 show RT (input user, partner user) for the cases where different users act as the input users.
RT(user A,user B)=3×1=3 (Expression 21)
RT(user A,user C)=1×1=1 (Expression 22)
RT(user B,user A)=0 (Expression 23)
RT(user B,user C)=0 (Expression 24)
RT(user C,user A)=1×1=1 (Expression 25)
RT(user C,user B)=0 (Expression 26)
Based on the values obtained from the above Expressions 21 through 26, RTv can be calculated as shown in the following Expression 27, and C (user A, user B) can be calculated as shown in the following Expression 28 using the result of the above Expression 11.
RTv=(3+1+0+0+1+0)/6=5/6=approximately 0.83 (Expression 27)
C(user A,user B)=(2×1+3×1)/1.0+(3×1)/0.83=8.6 (Expression 28)
Similarly, provided that users A and C are the input user and the partner user, respectively, the user interest degree C. (user A, user C) can be calculated using the following Expression 29.
C(user A,user C)=(1×1)/1.0+(1×1)/0.83=2.2 (Expression 29)
Provided that users B and A are the input user and the partner user, respectively, similar calculation leads to the user interest degree C. (user B, user A)=0. Furthermore, provided that users C and A are the input user and the partner user, respectively, similar calculation leads to the user interest degree C. (user C, user A)=2.2, which is the same as Expression 29.
As set forth above, in the present Embodiment 2, when a user takes an action corresponding to a recommended action, the user interest degree increases.
Thereafter, the communication level determination unit 51 refers to the preference similarities between users accumulated in the preference similarity accumulation unit 14, and obtains preference similarities between the input user and all other users (step S15). Step S15 is similar to step S4 of Embodiment 1 shown in
The communication level determination unit 51 then determines a communication level based on the user interest degrees calculated in step S14 and the preference similarities obtained in step S15 (step S16). Step S16 is similar to step S5 of Embodiment 1 shown in
As set forth above, in the present Embodiment 2 also, the operations of the communication assistance device 50 allow determining a communication level based on similarities between preference patterns of users and records of actions taken by a user toward other users, as in Embodiment 1. The communication assistance device 50 can provide communication assistance by determining the level in accordance with the level. Furthermore, the present Embodiment 2 takes into consideration whether or not a user has followed recommended actions in determining a communication level, and therefore improves the accuracy of the determination.
It suffices for a program according to the present Embodiment 2 to cause a computer to execute steps S11 to S16 shown in
The preference similarity accumulation unit 14 is realized by the CPU and a storage device, such as a hard disk, provided in the computer. Furthermore, the communication level accumulation unit 13, the user action record accumulation unit 15 and the recommended action record accumulation unit 52 are realized by the storage device.
The following describes a communication assistance device, a communication assistance method and a program according to Embodiment 3 of the present invention with reference to
As shown in
The topic recommendation unit 61 receives, as the input, the user name input to the input reception unit 12. Based on the received user name, the topic recommendation unit 61 refers to the topics accumulated in the topic accumulation unit 63 and the user models accumulated in the user model accumulation unit 64, and searches for topics to be recommended to the input user and the partner user.
More specifically, based on the topics accumulated in the topic accumulation unit 63 and the user models accumulated in the user model accumulation unit 64, the topic recommendation unit 61 first calculates a degree of preference for each topic on a per-user basis.
Then, in order to provide communication assistance to users at various communication levels (including both the input user and the partner user), the topic recommendation unit 61 narrows the group of topics accumulated in the topic accumulation unit 63 down to topic candidates that can be transmitted to the partner user. For example, when the partner user is at a friend level, the topic candidates are topics for which at least only the partner user exhibits a high degree of preference. When the partner user is at other levels, the topic candidates are topics for which both the input user and the partner user exhibit a high degree of preference.
This is because when users are at a friend level, they know each other well, and therefore it is preferable to transmit a topic that fits at least the preferences of a partner user to the partner user for viewing purposes, even if the other user is not interested in that topic. On the other hand, when users are at other levels, their relationship is in the midst of cultivation, and therefore it is preferable to transmit a topic that fits both the preferences of a partner user and the preferences of the other user to the partner user for viewing purposes.
Thereafter, the topic recommendation unit 61 determines whether topics in which the partner user is interested out of the topic candidates should be recommended to the input user or to the partner user. For example, the topic recommendation unit 61 can select a user to whom information can be easily transmitted in consideration of the speed of transmission of information to users, namely the number of people to whom certain information can be transmitted per unit time.
More specifically, in this case, the topic recommendation unit 61 first refers to the communication levels accumulated in the communication level accumulation unit 13 and the user action records accumulated in the user action record accumulation unit 15. The topic recommendation unit 61 then identifies a user with a high user interest degree, or a user who has taken a large number of transmission actions (e.g. message transmissions) as user actions, and determines the identified user as a recommendation target.
Alternatively, in consideration of contribution to a partner user with whom communication is performed, the topic recommendation unit 61 may determine a user who has made a small contribution to a partner user, namely a user with a low user interest degree, as the recommendation target. Note that “contribution to a partner user” denotes actions taken toward the partner user, and includes transmission of information to the partner user, transmission of return comments to the partner user, and the like. Contribution to a partner user represents actions necessary to maintain a human relationship. More specifically, in the present case, the topic recommendation unit 61 refers to the communication levels accumulated in the communication level accumulation unit, identifies a user with a low user interest degree, and determines the identified user as the recommendation target.
Furthermore, the topic recommendation unit 61 may change the above criterion for determining a user to be the recommendation target depending on communication levels. For example, in the case of the existence awareness level at the stranger level with similar preferences, the topic recommendation unit 61 may determine a user with a low user interest degree as the recommendation target in consideration of contribution to a partner user. This is because the existence awareness level is the state where the partner user is interested in the determined user and it is thus necessary for the determined user to voluntarily transmit information to move to the next level, i.e. the information exchange level.
Lastly, when a large number of topics have been determined to be recommended to the input user out of the topic candidates, the topic recommendation unit 61 adjusts the amount of topics to be recommended. For example, in the case where the aim of topic recommendation is to further advance an already-progressing communication level and form a deep human relationship, the topic recommendation unit 61 adopts a large number of topics for a partner user at a more advanced communication level. On the other hand, in the case where the aim of topic recommendation is to form broad human relationships with a large number of people, the topic recommendation unit 61 adopts a large number of topics for lower communication levels.
The topic accumulation unit 63 accumulates topics that serve as a medium of communication as shown in
The user model accumulation unit 64 accumulates information showing users' preferences, namely user models, as shown in
As shown in
In order to provide a user with topics to be recommended, the output unit 62 outputs the topics to the service providing device 30. The output unit 62 is not only capable of outputting the topics, but also capable of outputting an action for transmitting the output topics to a communication partner. Furthermore, when outputting such an action, the output unit 62 can instruct the service providing device 30 not to display the user name of the communication partner depending on communication levels.
More specifically, at a level closer to the stranger level than to the stranger level with similar preferences and no knowledge of name, i.e. the intermediate level, the output unit 62 instructs the service providing device 30 not to display the user name of the communication partner. In this case, the output unit 62 may request the service providing device 30 to display information that substitutes the user name of the communication partner. For example, in the case where a user to whom topics are provided can read user models, the above information may be information with the largest weight out of all pieces of preference information of a partner user. Alternatively, the above information may be a keyword that has been determined to have the largest weight through analysis of topic information, or a keyword for which the largest weight has been calculated from topics and a user model of a partner user.
A description is now given of the operations of the communication assistance device 60 according to Embodiment 3 of the present invention with reference to
As shown in
First, steps S21 to S26 are executed. Steps S21 to S26 are similar to steps S11 to S16 in
Thereafter, the topic recommendation unit 61 calculates a degree of preference for each topic on a per-user basis based on the topics accumulated in the topic accumulation unit 63 and the user models accumulated in the user model accumulation unit 64 (step S27). Note that a degree of preference denotes a degree at which a user likes a topic. The larger the value of the degree of preference is, the more the topic matches the user's preferences.
In step S27, a known method may be used to calculate a degree of preference. For example, the topic recommendation unit 61 first refers to the user models accumulated in the user model accumulation unit 64 and generates a keyword vector of each user model. The topic recommendation unit 61 then performs language analysis on the topics accumulated in the topic accumulation unit 63 and generates a topic keyword vector from the result of the analysis. Thereafter, the topic recommendation unit 61 calculates cosine similarity between the keyword vector of each user model and the topic keyword vector as a degree of preference. At this time, the topic recommendation unit 61 calculates the degrees of preference for all pairs of user and topic.
Next, the topic recommendation unit 13 narrows the topics down to topic candidates that can be transmitted to a partner user (step S28). As stated earlier, in step S28, the topic recommendation unit 61 selects topic candidates as follows. When a partner user is at a friend level, the topic recommendation unit 61 selects topics for which at least only the partner user exhibits a high degree of preference as topic candidates. When a partner user is at other levels, the topic recommendation unit 61 selects topics for which both the input user and the partner user exhibit a high degree of preference as topic candidates.
For example, assume that the communication levels accumulated in the communication level accumulation unit 13 are in the state shown in
For example, provided that a threshold for a degree of preference is set to 0.5, topic 1 and topic 2 are selected as topic candidates from among the result of the calculation shown in
Regarding topics in which a partner user is interested out of the topic candidates, the topic recommendation unit 61 determines a user to whom these topics are to be recommended as a recommendation target (step S29). For example, in the present example, it is assumed that the topic recommendation unit 61 determines a user to whom information can be easily transmitted as a recommendation target in consideration of the speed of information transmission (information transmissibility). In this case, the topic recommendation unit 61 refers to the communication levels accumulated in the communication level accumulation unit 13 and the user action records accumulated in the user action record accumulation unit 15. The topic recommendation unit 61 then determines a user with a high user interest degree, or a user who has taken a large number of transmission actions (e.g. message transmissions) as user actions, as a recommendation target.
For example, regarding topic 1, it is necessary to compare user A with user B and compare user A with user C. Referring to
When a large number of topics have been determined to be recommended to the input user out of the topic candidates, the topic recommendation unit 61 adjusts the amount of topics to be recommended (step S30). In the present example, it is assumed that the aim of the service providing device 30 is to form broad human relationships with a large number of people. In this case, the topic recommendation unit 61 adjusts the amount of topics to be recommended so that a larger number of topics are prepared for lower communication levels. More specifically, an upper limit is set for the number of topics for each level, so that a larger number of topics are adopted for lower communication levels. Topics are selected in order from a topic for which the input user or the partner user exhibits the highest degree of preference.
For example, assume that three is set as an upper limit for the number of topics for a low communication level, namely a stranger level, two is set as an upper limit for the number of topics for a stranger level with similar preferences and no knowledge of name, one is set as an upper limit for the number of topics for the next level, namely a candidate-for-friend level with similar preferences, and one is set as an upper limit for the number of topics for a candidate-for friend level. In this case, in the present example, topic 1 applies to both the candidate-for-friend level with similar preferences (user A, user B) and the stranger level with similar preferences and no knowledge of name (user A, user C). Therefore, topic 1 may be considered to apply to either one of these levels. On the other hand, topic 2 applies to the candidate-for-friend level with similar preferences (user A, user B). As the conditions regarding the aforementioned upper limits are satisfied, topic 1 and topic 2 are both adopted.
Thereafter, in order to provide a user with topics to be recommended, the output unit 62 outputs the topics to the service providing device 30 (step S31). In the above example, “topic 1 (user B, user C)” and “topic 2 (user B)” are output in step S31. Here, parentheses show partner users of a user who has been determined as a recommendation target. In this case, the service providing device 30 causes a display of the terminal device 20 of the user who has been determined as the recommendation target to display the topics and the user names of the partner users.
When outputting topics, the output unit 62 can also instruct the service providing device 30 to recommend an action for transmitting the output topics to the partner users. At this time, when a communication level between the recommendation target to whom the topics are output and his/her partner user is closer to the stranger level than to the stranger level with similar preferences and no knowledge of name, the output unit 62 instructs the service providing device 30 not to display the user name of the partner user.
The following describes an example of the output displayed on the display of the terminal device 20 of a user with reference to
As set forth above, with the communication assistance device 60 according to the Embodiment 3, the users are given a chance to communicate with other users. As a result, communication is facilitated. Furthermore, when the communication level is an initial level, namely non-advanced communication level, the user can hide own name and communicate with other users. As a result, user easily overcome psychological obstacles, communication is facilitated in this regard.
It suffices for a program according to the present Embodiment 3 to cause a computer to execute steps S21 to S31 shown in
The preference similarity accumulation unit 14 is realized by the CPU and a storage device, such as a hard disk, provided in the computer. Furthermore, the communication level accumulation unit 13, the user action record accumulation unit 15, the recommended action record accumulation unit 52, the topic accumulation unit 63 and the user model accumulation unit 64 are realized by the storage device.
With reference to
As shown in
The CPU 111 performs various types of arithmetic operations by deploying programs (codes) of the embodiments stored in the storage device 113 to the main memory 112 and executing the deployed programs in a predetermined order. The main memory 112 is typically a volatile storage device such as a dynamic random-access memory (DRAM). The programs of the embodiments are provided while being stored in a computer-readable recording medium 120. The programs of the embodiments may be distributed over the Internet connected via the communication interface 117.
Specific examples of the storage device 113 include a hard disk and a semiconductor storage device such as a flash memory. The input interface 114 mediates data transmission between the CPU 111 and an input device 118 such as a keyboard and a mouse. The display controller 115 is connected to the display device 119 and controls display on the display device 119. The data reader/writer 116 mediates data transmission between the CPU 111 and the recording medium 120. More specifically, the data reader/writer 116 reads programs from the recording medium 120 and writes the results of processing executed by the computer 110 to the recording medium 120. The communication interface 117 mediates data transmission between the CPU 111 and other computers.
Specific examples of the recording medium 120 include a general-purpose semiconductor storage device such as a CompactFlash (CF) and a Secure Digital (SD), a magnetic storage medium such as a flexible disk, and an optical storage medium such as a Compact Disc read-only memory (CD-ROM). Note that in Embodiment 1, Embodiment 2 and Embodiment 3, the computer 110 may be a server computer that constitutes the service providing device 30, or may be another computer.
A part or all of the above embodiments can be described as, but are not limited to, the following Notes 1 to 27.
(Note 1)
A communication assistance device including:
a communication level determination unit that determines a level of a relationship between users based on similarity between the users and on user action records, the similarity between the users being obtained from preference information showing preferences of the users, and the user action records showing records of actions taken by a certain user toward a partner user with whom the certain user communicates out of the users.
(Note 2)
The communication assistance device according to Note 1, wherein
the communication level determination unit determines whether the relationship between the users is at a stranger level, at an intermediate level between strangers and friends, or at a friend level.
(Note 3)
The communication assistance device according to Note 2, wherein
the intermediate level includes at least one of a stranger level with similar preferences and no knowledge of name, at which the certain user is similar to the partner user in preferences but is not interested in the partner user; and a candidate-for-friend level with similar preferences, at which the certain user is similar to the partner user in preferences and is interested in the partner user, and
when determining that the relationship between the users is at the intermediate level, the communication level determination unit determines whether the relationship between the users is at the stranger level with similar preferences and no knowledge of name or at the candidate-for-friend level with similar preferences.
(Note 4)
The communication assistance device according to Note 3, wherein
the stranger level with similar preferences and no knowledge of name and the candidate-for-friend level with similar preferences are each classified into: an existence awareness level at which the existence of the partner user is acknowledged; and an information exchange level at which information is exchanged with the partner user, and
when determining that the relationship between the users is at the intermediate level, the communication level determination unit determines whether the relationship between the users is at the existence awareness level of the stranger level with similar preferences and no knowledge of name, at the information exchange level of the stranger level with similar preferences and no knowledge of name, at the existence awareness level of the candidate-for-friend level with similar preferences, or at the information exchange level of the candidate-for-friend level with similar preferences.
(Note 5)
The communication assistance device according to any of Notes 1 to 4, wherein
the communication level determination unit
identifies, from the user action records, details of actions that have been taken by the certain user toward the partner user and the number of times the actions have been taken in a preset time period,
calculates a degree of interest that the certain user takes in the partner user based on the identified details of the actions and the identified number of times the actions have been taken, and
judges whether the relationship between the users is in an advancing state, in a stalled state, or in a receding state based on the similarity between the users and the degree of interest taken in the partner user, and determines the level of the relationship between the users using a result of the judgment.
(Note 6)
The communication assistance device according to any of Notes 1 to 5, wherein
the communication level determination unit
judges that the relationship between the users is in the advancing state when the similarity between the users is equal to or greater than a first predetermined value set therefor and the degree of interest taken in the partner user is equal to or greater than a first predetermined value set therefor, and
judges that the relationship between the users is in the receding state when the similarity between the users is equal to or smaller than a second predetermined value set therefor or when the degree of interest taken in the partner user is equal to or smaller than a second predetermined value set therefor.
(Note 7)
The communication assistance device according to any of Notes 1 to 6, wherein
in a case where an external device recommends actions to each user and each user takes the recommended actions,
the communication level determination unit
identifies, from the user action records, details of actions that have been taken by the certain user toward the partner user in a preset time period, the number of times the actions have been taken, details of the recommended actions taken by the certain user, and the number of times the recommended actions have been taken,
calculates a degree of interest that the certain user takes in the partner user based on the identified details of the actions, the identified number of times the actions have been taken, the identified details of the recommended actions, and the identified number of times the recommended actions have been taken, and
judges whether the relationship between the users is in an advancing state, in a stalled state, or in a receding state based on the similarity between the users and the degree of interest taken in the partner user, and determines the level of the relationship between the users using a result of the judgment.
(Note 8)
The communication assistance device according to any of Notes 1 to 7, further includes a topic recommendation unit that selects, from among a group of topics that have been prepared in advance, a topic that can be transmitted to the partner user based on the level of the relationship between the users determined by the communication level determination unit and on preferences of the certain user and the partner user.
(Note 9)
The communication assistance device according to Notes 8, wherein
the topic recommendation unit
calculates degrees of preference for each topic included among the group of topics on a per-user basis, and selects, as the topic that can be transmitted to the partner user, a topic for which only the partner user exhibits a degree of preference equal to or greater than a set value, or a topic for which both the certain user and the partner user exhibit a degree of preference equal to or greater than the set value, based on the level of the relationship between the users determined by the communication level determination unit and on the degrees of preference calculated on the per-user basis,
determines a recommendation target user to whom the selected topic is to be recommended out of the certain user and the partner user based on information transmissibility of the certain user and the partner user or on contribution that the certain user and the partner user make to each other, and
further determines the number of topics to be recommended to the recommendation target user in accordance with the level of the relationship between the users.
(Note 10)
A communication assistance method including:
a communication level determination step of determining a level of a relationship between users based on similarity between the users and on user action records, the similarity between the users being obtained from preference information showing preferences of the users, and the user action records showing records of actions taken by a certain user toward a partner user with whom the certain user communicates out of the users.
(Note 11)
The communication assistance method according to Note 10, wherein
the communication level determination step determines whether the relationship between the users is at a stranger level, at an intermediate level between strangers and friends, or at a friend level.
(Note 12)
The communication assistance method according to Note 11, wherein
the intermediate level includes at least one of a stranger level with similar preferences and no knowledge of name, at which the certain user is similar to the partner user in preferences but is not interested in the partner user; and a candidate-for-friend level with similar preferences, at which the certain user is similar to the partner user in preferences and is interested in the partner user, and
when determining that the relationship between the users is at the intermediate level, the communication level determination step determines whether the relationship between the users is at the stranger level with similar preferences and no knowledge of name or at the candidate-for-friend level with similar preferences.
(Note 13)
The communication assistance method according to Note 12, wherein
the stranger level with similar preferences and no knowledge of name and the candidate-for-friend level with similar preferences are each classified into: an existence awareness level at which the existence of the partner user is acknowledged; and an information exchange level at which information is exchanged with the partner user, and
when determining that the relationship between the users is at the intermediate level, the communication level determination step determines whether the relationship between the users is at the existence awareness level of the stranger level with similar preferences and no knowledge of name, at the information exchange level of the stranger level with similar preferences and no knowledge of name, at the existence awareness level of the candidate-for-friend level with similar preferences, or at the information exchange level of the candidate-for-friend level with similar preferences.
(Note 14)
The communication assistance method according to any of Notes 10 to 13, wherein
the communication level determination step
identifies, from the user action records, details of actions that have been taken by the certain user toward the partner user and the number of times the actions have been taken in a preset time period,
calculates a degree of interest that the certain user takes in the partner user based on the identified details of the actions and the identified number of times the actions have been taken, and
judges whether the relationship between the users is in an advancing state, in a stalled state, or in a receding state based on the similarity between the users and the degree of interest taken in the partner user, and determines the level of the relationship between the users using a result of the judgment.
(Note 15)
The communication assistance method according to any of Notes 10 to 14, wherein
the communication level determination step
judges that the relationship between the users is in the advancing state when the similarity between the users is equal to or greater than a first predetermined value set therefor and the degree of interest taken in the partner user is equal to or greater than a first predetermined value set therefor, and
judges that the relationship between the users is in the receding state when the similarity between the users is equal to or smaller than a second predetermined value set therefor or when the degree of interest taken in the partner user is equal to or smaller than a second predetermined value set therefor.
(Note 16)
The communication assistance method according to any of Notes 10 to 15, wherein
in a case where an external device recommends actions to each user and each user takes the recommended actions,
the communication level determination step
identifies, from the user action records, details of actions that have been taken by the certain user toward the partner user in a preset time period, the number of times the actions have been taken, details of the recommended actions taken by the certain user, and the number of times the recommended actions have been taken,
calculates a degree of interest that the certain user takes in the partner user based on the identified details of the actions, the identified number of times the actions have been taken, the identified details of the recommended actions, and the identified number of times the recommended actions have been taken, and
judges whether the relationship between the users is in an advancing state, in a stalled state, or in a receding state based on the similarity between the users and the degree of interest taken in the partner user, and determines the level of the relationship between the users using a result of the judgment.
(Note 17)
The communication assistance method according to any of Notes 10 to 16, further includes a topic recommendation step of selecting, from among a group of topics that have been prepared in advance, a topic that can be transmitted to the partner user based on the level of the relationship between the users determined in the communication level determination step and on preferences of the certain user and the partner user.
(Note 18)
The communication assistance method according to Note 17, wherein
the topic recommendation step
calculates degrees of preference for each topic included among the group of topics on a per-user basis, and selects, as the topic that can be transmitted to the partner user, a topic for which only the partner user exhibits a degree of preference equal to or greater than a set value, or a topic for which both the certain user and the partner user exhibit a degree of preference equal to or greater than the set value, based on the level of the relationship between the users determined by the communication level determination unit and on the degrees of preference calculated on the per-user basis,
determines a recommendation target user to whom the selected topic is to be recommended out of the certain user and the partner user based on information transmissibility of the certain user and the partner user or on contribution that the certain user and the partner user make to each other, and
further determines the number of topics to be recommended to the recommendation target user in accordance with the level of the relationship between the users.
(Note 19)
A computer-readable recording medium having recorded therein a program including instructions for causing a computer to execute:
a communication level determination step of determining a level of a relationship between users based on similarity between the users and on user action records, the similarity between the users being obtained from preference information showing preferences of the users, and the user action records showing records of actions taken by a certain user toward a partner user with whom the certain user communicates out of the users.
(Note 20)
The computer-readable recording medium according to Note 19, wherein
the communication level determination step determines whether the relationship between the users is at a stranger level, at an intermediate level between strangers and friends, or at a friend level.
(Note 21)
The computer-readable recording medium according to Note 20, wherein
the intermediate level includes at least one of a stranger level with similar preferences and no knowledge of name, at which the certain user is similar to the partner user in preferences but is not interested in the partner user; and a candidate-for-friend level with similar preferences, at which the certain user is similar to the partner user in preferences and is interested in the partner user, and
when determining that the relationship between the users is at the intermediate level, the communication level determination step determines whether the relationship between the users is at the stranger level with similar preferences and no knowledge of name or at the candidate-for-friend level with similar preferences.
(Note 22)
The computer-readable recording medium according to Note 21, wherein
the stranger level with similar preferences and no knowledge of name and the candidate-for-friend level with similar preferences are each classified into: an existence awareness level at which the existence of the partner user is acknowledged; and an information exchange level at which information is exchanged with the partner user, and
when determining that the relationship between the users is at the intermediate level, the communication level determination step determines whether the relationship between the users is at the existence awareness level of the stranger level with similar preferences and no knowledge of name, at the information exchange level of the stranger level with similar preferences and no knowledge of name, at the existence awareness level of the candidate-for-friend level with similar preferences, or at the information exchange level of the candidate-for-friend level with similar preferences.
(Note 23)
The computer-readable recording medium according to any of Notes 19 to 22, wherein
the communication level determination step
identifies, from the user action records, details of actions that have been taken by the certain user toward the partner user and the number of times the actions have been taken in a preset time period,
calculates a degree of interest that the certain user takes in the partner user based on the identified details of the actions and the identified number of times the actions have been taken, and
judges whether the relationship between the users is in an advancing state, in a stalled state, or in a receding state based on the similarity between the users and the degree of interest taken in the partner user, and determines the level of the relationship between the users using a result of the judgment.
(Note 24)
The computer-readable recording medium according to any of Notes 19 to 23, wherein
the communication level determination step
judges that the relationship between the users is in the advancing state when the similarity between the users is equal to or greater than a first predetermined value set therefor and the degree of interest taken in the partner user is equal to or greater than a first predetermined value set therefor, and
judges that the relationship between the users is in the receding state when the similarity between the users is equal to or smaller than a second predetermined value set therefor or when the degree of interest taken in the partner user is equal to or smaller than a second predetermined value set therefor.
(Note 25)
The computer-readable recording medium according to any of Notes 19 to 24, wherein
in a case where an external device recommends actions to each user and each user takes the recommended actions,
the communication level determination step
identifies, from the user action records, details of actions that have been taken by the certain user toward the partner user in a preset time period, the number of times the actions have been taken, details of the recommended actions taken by the certain user, and the number of times the recommended actions have been taken,
calculates a degree of interest that the certain user takes in the partner user based on the identified details of the actions, the identified number of times the actions have been taken, the identified details of the recommended actions, and the identified number of times the recommended actions have been taken, and
judges whether the relationship between the users is in an advancing state, in a stalled state, or in a receding state based on the similarity between the users and the degree of interest taken in the partner user, and determines the level of the relationship between the users using a result of the judgment.
(Note 26)
In the computer-readable recording medium according to any of Notes 19 to 25, the program further includes a topic recommendation step of selecting, from among a group of topics that have been prepared in advance, a topic that can be transmitted to the partner user based on the level of the relationship between the users determined in the communication level determination step and on preferences of the certain user and the partner user.
(Note 27)
The computer-readable recording medium according to Note 26, wherein
the topic recommendation step
calculates degrees of preference for each topic included among the group of topics on a per-user basis, and selects, as the topic that can be transmitted to the partner user, a topic for which only the partner user exhibits a degree of preference equal to or greater than a set value, or a topic for which both the certain user and the partner user exhibit a degree of preference equal to or greater than the set value, based on the level of the relationship between the users determined by the communication level determination unit and on the degrees of preference calculated on the per-user basis,
determines a recommendation target user to whom the selected topic is to be recommended out of the certain user and the partner user based on information transmissibility of the certain user and the partner user or on contribution that the certain user and the partner user make to each other, and
further determines the number of topics to be recommended to the recommendation target user in accordance with the level of the relationship between the users.
Although the invention of the present application has been described using the above embodiments, the invention of the present application is by no means limited to the above embodiments. The configurations and details of the invention of the present application may be subjected to various changes that can be understood by a person skilled in the art within a scope of the invention of the present application.
The present application claims the benefit of priority from Japanese Patent Application No. 2010-177374, filed Aug. 6, 2010, the disclosure of which is incorporated herein by reference in its entirety.
As set forth above, the present invention allows determining a level of a relationship between users who communicate with each other. Therefore, the present invention can be applied to, for example, a conversation assistance device that recommends a conversation partner, and a program for realizing the conversation assistance device using a computer.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2010-177374 | Aug 2010 | JP | national |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/JP2011/065960 | 7/13/2011 | WO | 00 | 1/16/2013 |
