JP2015056045A - Information processing device, information processing method and program, and user terminal device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To propose content information suitable of a user.SOLUTION: An information recommendation response processing unit 220 is configured to: receive a recommendation request for content information from a mobile terminal 100 via a communication unit 210 to search for the content information matching a condition; calculate similarity between a departure point, a destination and a transmission destination of the recommendation request when other users selecting the retrieved content information in the past selects the content information and a departure point of a user using the mobile terminal 100, a destination thereof and the transmission destination of the recommendation request; as to combinations of registered users, calculate similarity about distances between activity bases of users pertain to the combinations thereof and destinations common among the users pertain to the combinations thereof; and calculate an evaluation value of the content information using the calculated two kinds of similarity, and transmit the content information and a recommendation result having the evaluation value of the content information indicated to the mobile terminal 100 via the communication part 210.

Description

  The present invention, for example, is valuable for the user from content information (hereinafter also simply referred to as content) such as a myriad of movement candidate points based on position information obtained from a mobile terminal such as a car navigation system or a smartphone. The present invention relates to a technique for recommending certain content information.

In the conventional information recommendation method, the behavior of a user who is acting away from the user's activity area (living area) is detected, and the visit history of other users who have visited the area away from the activity area is also recommended. By using it as a candidate for an object, in a region where the user is unfamiliar, the recommendation of information suitable for a traveler is performed instead of the recommendation of a local person (for example, Patent Document 1).
In addition, in the method using collaborative filtering, which is a conventional representative method for recommending content information that matches the user's preference, between content information calculated by the same evaluation from multiple users at the same time There is a method of performing content evaluation and information recommendation suitable for the user's preference by evaluating the content information using the similarity degree of the user and the past content evaluation information of the user (for example, Patent Document 2).

International Publication WO2004 / 075137 International Publication No. WO2011 / 074113

The method according to the conventional patent document 1 recommends a point where another user has visited many times away from the active area, or at least the time, stay time, day of the week, weather, and season when the other user visits. One is to recommend points that match the current situation.
For this reason, it cannot be determined whether the recommended spot is suitable for the preference of the user receiving the recommendation, and there is a problem that it may not be a valuable candidate spot for a specific user.

  Moreover, since the method according to Patent Document 2 considers only the past evaluation relationship of the user, there is a problem that a candidate point suitable for the current user situation cannot be recommended.

  The main object of the present invention is to solve the above-mentioned problems, and the main object is to present content information suitable for the user.

An information processing apparatus according to the present invention includes:
A search request indicating a start location, a destination, and a search request transmission location of a user moving from the start location to the destination is received from the user terminal device used by the user, and the destination and search indicated in the search request are received. An information processing device that searches content information related to at least one of request transmission locations and transmits a search response that notifies the searched content information to the user terminal device,
User base information indicating activity bases of a plurality of users, search request history information indicating the contents of a search request from a search request history user who has transmitted a search request among the plurality of users, A storage for storing selection history information indicating content information selected by a selection history user who has selected any content information from among the plurality of users notified by a search response. And
A content information search unit for searching for content information related to at least one of the destination and the search request transmission location indicated in the reception search request received from the user terminal device of any one of the plurality of users;
Based on the search request history information, the selection history information, and the received search request, the target selection history user who has selected the search content information searched by the content information search unit and the request source of the received search request A common destination with a requesting user, a distance between the common destination and the activity base of the requesting user indicated in the user base information, and the common destination and the user base information. A content information evaluation value calculation unit that calculates an evaluation value of the search content information using a distance difference between the target selection history user and the distance to the activity base shown;
It has a search response transmission part which transmits the search response which notifies the said search content information and the evaluation value calculated by the said content information evaluation value calculation part to the said user terminal device, It is characterized by the above-mentioned.

  According to the present invention, the current situation of the requesting user, the situation when another user has selected the content information, the distance from the activity base of the requesting user to the common destination, and the activity base of the other user In order to calculate the evaluation value of the content information using the distance from and present the content information to the requesting user together with the evaluation value, the appropriate content information is considered in consideration of the current situation and action pattern of the requesting user. Can be presented.

1 is a diagram illustrating an example of the overall configuration of an information system according to Embodiment 1. FIG. FIG. 3 is a diagram illustrating a detailed configuration example of the information recommendation device according to the first embodiment. FIG. 4 is a flowchart showing an example of a processing flow of an information recommendation request processing unit according to the first embodiment. FIG. 4 is a flowchart showing an example of a processing flow of an information recommendation response processing unit according to the first embodiment. FIG. 4 is a flowchart showing an example of a processing flow of an information recommendation response processing unit according to the first embodiment. FIG. 3 is a flowchart showing an example of a content evaluation processing flow according to the first embodiment. The flowchart figure which shows the example of a processing flow of the preference similarity calculation classified by distance which concerns on Embodiment 1. FIG. FIG. 4 is a flowchart showing a processing flow example of situation similarity calculation according to the first embodiment. FIG. 4 is a diagram showing an example of a user information management table according to the first embodiment. FIG. 4 is a diagram showing an example of a content information management table according to the first embodiment. FIG. 4 is a diagram showing an example of a search history management table according to the first embodiment. FIG. 4 is a diagram showing an example of a recommendation result management table according to the first embodiment. FIG. 3 is a diagram illustrating a hardware configuration example of the information recommendation device according to the first embodiment.

Embodiment 1 FIG.
In the present embodiment, a configuration that enables the recommendation of content information (for example, candidate points) suitable for the user's preference and the user's current situation even in a region away from the user's active area. explain.

FIG. 1 is an example of an overall configuration diagram of the information system according to the present embodiment.
In FIG. 1, the information system according to the present embodiment includes a mobile terminal 100, an information recommendation device 200, and a network 300 that is an environment for communicating between them.

The mobile terminal 100 is a terminal device for requesting information recommendation from the information recommendation device 200.
There may be a plurality of mobile terminals 100.
The mobile terminal 100 is an example of a user terminal device.
The mobile terminal 100 includes a communication unit 110, an information recommendation request processing unit 120, a position information acquisition unit 130, and a navigation processing unit 140.

The communication unit 110 transmits a recommendation request (an example of a search request) for requesting search / recommendation of content information to the information recommendation device 200, and receives a response from the information recommendation device 200.
The communication unit 110 may be implemented with a general communication protocol, such as TCP (Transmission Control Protocol).

The position information acquisition unit 130 acquires the position of the mobile terminal 100.
The position information acquisition unit 130 may be a general position acquisition unit, such as a GPS (Global Positioning System).

The information recommendation request processing unit 120 requests the information recommendation response processing unit 220 of the information recommendation device 200 to search and recommend content information.
In the present embodiment, the information recommendation request processing unit 120 has a function of accepting user operations such as search, display of recommendation results, display of content information, and selection of recommendation results.

The navigation processing unit 140 performs navigation of a vehicle or the like.
Here, an example in which the navigation processing unit 140 is used has been described, but not limited to the navigation processing unit 140, any means (program) that utilizes the recommendation result may be installed in the mobile terminal 100.

The information recommendation device 200 responds to the information recommendation request from the mobile terminal 100.
The information recommendation device 200 is an example of an information processing device.
The information recommendation device 200 includes a communication unit 210, an information recommendation response processing unit 220, and a recommendation information storage unit 230.

The communication unit 210 receives a recommendation request from the mobile terminal 100, and transmits a recommendation result (an example of a search response) that is a response to the recommendation request to the mobile terminal 100.
Similar to the communication unit 110, the communication unit 210 may be implemented with a general communication protocol, such as TCP.

  The information recommendation response processing unit 220 receives the recommendation request from the information recommendation request processing unit 120, evaluates the content information to be recommended, and generates a recommendation result for notifying the content information to be recommended and the evaluation value of the content information To do.

The recommendation information storage unit 230 stores data used by the information recommendation response processing unit 220 when evaluating content information.
The recommendation information storage unit 230 may be implemented by a database such as an RDB (Relational Database).

The network 300 is a network for communication between the mobile terminal 100 and the information recommendation device 200.
The network 300 may be the Internet.

  FIG. 2 shows a detailed configuration example of the information recommendation device 200.

In FIG. 2, as described above, the communication unit 210 receives a recommendation request from the mobile terminal 100 and transmits a recommendation result that is a response to the recommendation request to the mobile terminal 100.
The communication unit 210 is an example of a search request receiving unit and a search response transmitting unit.
The distance preference similarity calculation unit 221 calculates the preference similarity of the distance from the current position among all users to the target position.
The distance preference similarity calculation unit 221 is an example of a distance similarity calculation unit.

  The situation similarity calculation unit 222 calculates the degree of situation similarity between the current situation of the user to be recommended (recommendation request requesting user) and the search situation of other users in the past.

The similarity propagation calculation unit 223 indirectly calculates the similarity between users using the similarity with other users capable of calculating the similarity when the direct similarity between users cannot be calculated. To do.
The similarity propagation calculation unit 223 is an example of a content information evaluation value calculation unit.

The preference / situation-considering content evaluation unit 224 searches for recommended candidate content information based on the recommendation request, and calculates an evaluation value from the similarity of the preference / situation for the searched content information.
The preference / situation-considering content evaluation unit 224 is an example of a content information search unit and a content information evaluation value calculation unit.

  The recommendation information management unit 225 manages a user search history and a recommendation result.

The information recommendation response unit 226 generates a recommendation result and transmits the recommendation result to the mobile terminal 100 via the communication unit 210.
In addition, the information recommendation response unit 226 performs control in the information recommendation device 200 from reception of a recommendation request to transmission of a recommendation result.

  The recommendation information storage unit 230 stores a user information management table 231, a content information management table 232, a search history management table 233, and a recommendation result management table 234.

The user information management table 231 is a table for managing information of users who have recommended information in the past and users who will be recommended information from now on.
As will be described later, the user information management table 231 is information indicating the activity base of each user, and corresponds to an example of user base information.

  The content information management table 232 is a table for managing content information recommended by information recommendation.

The search history management table 233 is a table for managing inquiry contents of information recommendation from the user.
As will be described later, the search history management table 233 includes a user who has transmitted a recommendation request (search request history user) among a plurality of users, and a departure place and purpose indicated in the recommendation request transmitted by the user. This is information in which the location and the transmission location of the recommendation request are shown in association with each other, and corresponds to an example of search request history information.

The recommendation result management table 234 is a table for managing the recommendation result and the content information selected by the user with respect to the recommendation result.
As will be described later, the recommendation result management table 234 includes a user (selection history user) who has selected any content information from among content information notified by the recommendation result among a plurality of users, and the user. This is information that is associated with the selected content information and corresponds to an example of selection history information.

  As described above, the recommendation information storage unit 230 includes a user information management table 231 that is an example of user base information, a search history management table 233 that is an example of search request history information, and a recommendation result management that is an example of selection history information. The table 234 is stored and corresponds to an example of a user base information storage unit, a search request history information storage unit, and a selection history information storage unit.

Next, the operation will be described.
FIG. 3 is a flowchart showing an example of processing of the information recommendation request processing unit 120 of the mobile terminal 100.

S100 shows an example in which the user uses the information recommendation request processing unit 120 and the navigation processing unit 140 using the mobile terminal 100.
Here, the mobile terminal 100 is assumed to be a car navigation system, and the flow from when the user starts searching until the recommended content is selected and navigation is started is shown as an example.
As an example of the scenario, it is assumed that a user “Mitsubishi Taro” living in Kamakura City, Kanagawa Prefecture, performs a search in the vicinity of Hakone when traveling to Nagoya City, Aichi Prefecture on a trip.
It is assumed that the user information management table 231 includes contents as shown in FIG.
In FIG. 9, 511 indicates a user ID, 512 indicates the name of the user, and 513 indicates the position of the home which is the user's activity base.
Here, for simplicity, the home is held by the city name, but it may be managed by coordinates and coordinate areas.
Further, it is assumed that the content information management table 232 includes contents as shown in FIG.
In FIG. 10, 521 indicates a content ID, 522 indicates the name of content information, 513 indicates a position, and 514 indicates a genre.
If there is information necessary for providing content information to the user, other items such as 515 may be added.
The user information management table 231 and the content information management table 232 may be registered in advance or may be obtained from outside the apparatus.

In S110, information recommendation request processing unit 120 accepts input of search information by an operation from the user.
As the search information, for example, a genre or keyword to be searched, a search condition, etc. are accepted.
These may be manually input by a user using a GUI (Graphical User Interface) or the like, or input by voice using a voice operation interface or the like.
Here, it is assumed that the user selects the genre of “restaurant” and “within 50 km” as a search condition.

In S120, the information recommendation request processing unit 120 acquires the situation where the search has been performed.
As a search situation, the user ID, the position information acquired by the position information acquisition unit 130, the car navigation processing unit 140 input by the user, or a preset starting point or destination are acquired.
Here, it is assumed that the user has set “35.233624, 139.105759” (near Hakone) as the position information, “Ofune” as the departure point, and “Nagoya” as the destination.
The user ID is “TARO”.

In S 130, information recommendation request processing unit 120 transmits a recommendation request to information recommendation response processing unit 220 via communication unit 110.
The recommendation request includes the search information acquired in S110 and S120 and the search status.

In S140, the information recommendation request processing unit 120 receives the recommendation result from the information recommendation response processing unit 220 via the communication unit 110, and displays the received recommendation result.
The recommendation result includes a recommendation result ID, a recommended content ID, and an evaluation value of each content.
Here, the recommendation result ID is “R001”, the recommended content IDs are “C001”, “C002”, and “C003”, and the evaluation values are “0.93”, “0.53”, and “0.48”, respectively. Suppose you get it.
The display order of the recommended content IDs may be sorted in the order of evaluation, or may be freely displayed such as by name or by keyword.

In S150, information recommendation request processing unit 120 accepts the content selected by the user from the recommendation results by an operation from the user.
Here, it is assumed that the user has selected “C001”.

In S160, the information recommendation request processing unit 120 sets the position of the selected content (position of the restaurant) in the car navigation processing unit 140 as a transit point.
As a result, “35.233534, 139.106994”, which is the position information of “C001”, is set at the waypoint on the car navigation.

In S170, the information recommendation request processing unit 120 transmits the ID of the content selected by the user to the information recommendation response processing unit 220.
The contents to be transmitted are the recommendation result ID extracted from the recommendation result received in S140 and the content ID selected in S150.
Here, “R001” is transmitted as the recommendation result ID, and “C001” is transmitted as the content ID.

  In S180, the car navigation processing unit 140 starts navigation toward the selected content (restaurant).

  4 and 5 are flowcharts showing an example of processing of the information recommendation response processing unit 220 of the information recommendation device 200. FIG.

S200 in FIG. 4 is a flow started upon receiving a recommendation request in S130 of the information recommendation request processing unit 120.
S300 in FIG. 5 is a flow started upon receiving the selection result in S170 of the information recommendation request processing unit 120.

  As shown in FIG. 4, in S <b> 210, the information recommendation response unit 226 receives a recommendation request from the information recommendation request processing unit 120 via the communication unit 210.

In S220, the information recommendation response unit 226 extracts the user ID, the departure place, the destination, the search point (the point where the recommendation request was transmitted), and the search information from the recommendation request received in S210, and extracts these elements. Stored in the search history management table 233.
Here, the user ID is “TARO”, the departure place is “Ofune”, the search points are “35.233624, 139.105759” (near Hakone), the destination is “Nagoya”, the genre is “restaurant”, and the search conditions Is “within 50 km”.
When making a recommendation using the travel route, the travel route from the departure point to the search point and the planned travel route from the search point to the destination are extracted in the same manner.

FIG. 11 shows the schema of the search history management table 233.
In FIG. 11, 531 is a column for storing a search ID, and the search history management table 233 automatically assigns a number.
Reference numeral 532 denotes a column for storing user IDs, which stores user IDs extracted from recommendation requests.
Reference numeral 533 denotes a column for storing departure points, which stores the names of departure points extracted from recommendation requests.
Reference numeral 534 denotes a column for storing destination points, which stores the name of the destination extracted from the recommendation request.
Reference numeral 535 denotes a column for storing search points, which stores the names of search locations extracted from recommendation requests.
Reference numeral 536 denotes a column for storing search conditions, which stores the search conditions extracted from the recommendation request.
Although not shown in FIG. 11, a column for storing the genre extracted from the recommendation request may be provided.
When the travel route is used for matching, a column for storing a travel route from the departure point to the search point and a planned travel route from the search point to the destination is provided.
Here, each point is managed by name, but if there is an identifier that can be shared with the navigation processing unit 140, it may be used or may be a coordinate.

In S230, the preference / situation-considering content evaluation unit 224 searches for content information that matches the conditions described in the recommendation request as a recommendation candidate, and calculates an evaluation value of the content information of the recommendation candidate.
This procedure will be described in detail in S400 described later.

In S240, the information recommendation response processing unit 220 receives the evaluation result performed in S230, and generates recommendation results in descending order of evaluation value.
For example, the recommendation result ID is “R001”, the recommended content IDs are “C001”, “C002”, “C003”, and the evaluation values are “0.93”, “0.53”, “0.48”, respectively. "Is generated according to the transmission format of the recommendation result.

  In S250, the information recommendation response unit 226 transmits the recommendation result generated in S240 to the information recommendation request processing unit 120, which is a request source, via the communication unit 210.

  As illustrated in FIG. 5, in S310, the recommended information management unit 225 receives the selection result in S170 from the information recommendation request processing unit 120.

In S320, the recommendation information management unit 225 extracts the recommendation result ID and the selected content ID from the selection result for the recommendation received in S310, and stores them in the recommendation result management table 234 shown in FIG.
Here, “R001” is stored as the recommendation result ID, and “C001” is stored as the selected content ID.
The recommendation result management table 234 may store a search ID, a recommended content ID, and an evaluation value.

  S400 of FIG. 6 is a detailed processing flow of content evaluation, which is a subflow of S230 of the information recommendation response processing unit 220.

In S410, the preference / situation considering content evaluation unit 224 extracts the activity bases of all users.
The preference / situation considering content evaluation unit 224 reads, for example, the user information management table 231 (FIG. 9) from the recommended information storage unit 230 and extracts home information registered in the user information management table 231 as an activity base.
Alternatively, the center of the past unit time is calculated from the search history management table 233 (FIG. 11) based on information such as the departure point and the destination point, and the calculated center is registered in the user information management table 231. The center may be extracted from the user information management table 231 as an activity base.
For example, the center is calculated as follows.
It is assumed that the departure point of the search history management table 233 is managed by coordinates.
From the search history management table 233, a departure point group for a user for the past month is acquired.
It is assumed that N starting points are obtained at this time.
The X coordinate and Y coordinate of each starting point are summed and divided by N, respectively.
By calculating as described above, it is possible to obtain the center where the user frequently operates from the search history.

In S420, the preference similarity by distance calculation unit 221 calculates the preference similarity by distance between all users.
This procedure will be described in detail in S421 described later.

In S430, the preference / situation consideration content evaluation unit 224 extracts a content information group to be recommended.
For example, if the “restaurant” genre and “within 50 km” are specified as search conditions, all content IDs that match the genre and search conditions are extracted from the content information management table 232 (FIG. 11).

In S440, the preference / situation-considering content evaluation unit 224 extracts users who have previously selected the content information extracted in S430.
More specifically, the preference / situation-considering content evaluation unit 224 reads the recommendation result management table 234 (FIG. 12) from the recommendation information storage unit 230 and selects the content ID acquired in S430 from the recommendation result management table 234. A search ID that has been selected in the past is extracted, and a user ID corresponding to the search ID is extracted from the search history management table 233 (FIG. 11).

In S450, the situation similarity calculation unit 222 calculates the situation similarity between the current situation of the user to be recommended (the user who requested the recommendation request) and the past search situation of other users.
This procedure will be described in detail in S451 described later.

In S460, the preference / situation-considering content evaluation unit 224 calculates the evaluation value of the content information from the preference / situation similarity.
The preference / situation considering content evaluation unit 224 calculates a combined similarity V (u, v) obtained by combining the preference similarity classified by distance calculated in S420 and the situation similarity calculated in S450.
u represents the requesting user of the recommendation request, and v represents the user (target selection history user) who has previously selected the content information (search content information) extracted in S430.
The combined similarity V (u, v) is the similarity between the requesting user u of the recommendation request and the user v who has previously selected the content information (search content information) extracted in S430, and the similarity between users Corresponds to degrees.
The combined similarity V (u, v) is calculated by the following equation.
In the following equation, Su is the preference similarity by distance calculated in S420, Ss is the situation similarity calculated in S450, and α and β are adjustment parameters for determining the respective weights.

Next, the preference / situation-considering content evaluation unit 224 calculates an evaluation value of the content information using the composite similarity V (u, v).
For example, the request source user of the recommendation request is “TARO”, the content information extracted in S430 is the content information A, and the users who have selected the content information A in the past are “JIRO” and “SABU”. Assume a case.
In this case, the combined similarity V between the user “TARO” and the user “JIRO” is expressed as V (T, J), and the combined similarity V between the user “TARO” and the user “SABU” is expressed as V. (T, S), the evaluation value of the content information A is represented as VA (T), and the situation similarity Ss between the user “TARO” and the user “JIRO” for the content information A is represented by SsA (T, J ) And the status similarity Ss between the user “TARO” and the user “JIRO” for the content information A is expressed as SsA (T, J), and the evaluation value of the content information for the user “TARO” VA (T) is calculated as follows.
VA (T) = {V (T, J) × SsA (T, J)} + {V (T, S) × SsA (T, S)}
In this way, a value obtained by multiplying the composite similarity V between users and the situation similarity Ss for each content is added to all the users who have selected the content in the past to obtain an evaluation value of the content information.
Content information selected in the past by another user who has a high degree of composite similarity with the requesting user of the recommendation request has a high evaluation value.

  S420 in FIG. 7 is a detailed processing flow for calculating the preference similarity by distance, which is a sub-flow of S420 in the processing flow S400 for content evaluation.

  In S421 and S422, the preference similarity calculation unit by distance 221 performs the processing from S423 onward for all combinations of users described in the user information management table 231 (FIG. 9).

For example, if the user ID “TARO” is selected in S421, the user IDs “JIRO”, “SABU”,... Are selected in S422, and the user ID “TARO” and the user IDs “JIRO”, “ The preference similarity according to distance with each of SABU "... is calculated.
In the following description, it is assumed that “TARO” is selected in S421 and “JIRO” is selected in S422.
The similarity calculation process between the same users may not be performed.

In S423, the preference similarity calculation unit 221 reads the search history management table 233 (FIG. 11) from the recommended information storage unit 230, and acquires a destination group shared by both from the search history management table 233.
That is, the preference similarity calculation unit 221 extracts a destination group that is common between “TARO” and “JIRO” from the search history management table 233.

In S424, the process of S425 is performed for all destinations on the destination group acquired in S423.
Here, it is assumed that “Hamamatsu Station” (coordinates Xh = 34.703972, Yh = 137.735405) is selected as the destination.

In S425, the preference similarity calculation unit 221 by distance calculates the distance between the activity base of each user and the destination, and obtains the difference.
For example, the distance from “Kamakura City” (coordinate Xk = 35.319242, Yk = 139.554601) of the activity base of the user “TARO” to “Hamamatsu Station” and “Fujisawa City” of the activity base of the user “JIRO” The distance from “coordinates: Xf = 35.338907, Yf139.490914” to “Hamamatsu Station” is obtained.
The distance Lkh between “Kamakura City” and “Hamamatsu Station” can be obtained from the three square theorem using the following formula.

Similarly, the distance Lfh between “Fujisawa City” and “Hamamatsu Station” can be obtained.
And the difference Ls of each distance with respect to the destination can be calculated | required as follows.

  Finally, the similarity St between users for each destination is obtained as follows by applying an evaluation function F (l) in which the evaluation value increases as the difference decreases.

ここで、Ｆ（ｌ）は以下のようなポアソン分布の累積分布関数を変形させた式などにて表す。

Here, F (l) is expressed by an equation obtained by modifying the cumulative distribution function of Poisson distribution as follows.

  As described above, the calculation of the similarity St according to the distance for each destination is repeated from step S424, and the total value of the similarities St for all the destinations is calculated.

In S426, the preference similarity calculation unit 221 by distance divides the value obtained by dividing the similarity by destination St obtained in S424 and S425 by the number of destinations shared by each user pair. Degree.
That is, the similarity Su (u, v) between the users u and v is calculated by the following equation.
Here, Tnum (u, v) is a function for calculating the number of destinations shared by the users u and v.

  S450 of FIG. 8 is a detailed processing flow of situation similarity calculation, which is a subflow of S450 in the content evaluation processing flow S400.

In S451, the situation similarity calculation unit 222 calculates the proximity between the search position and the past search position.
For example, Xt = 35.233624, Yt = 139.105759, which is the current position (recommendation request transmission point) of the user “TARO”, and Atami, which is the past search position (recommendation request transmission point) of the user “JIRO” The proximity to (coordinates Xj = 35.095879, Yj = 139.074478) is obtained by the following equation.
In the following equation, R ₁ is a function that gives higher evaluation as the distance is shorter.
For example, 1.0 is given if the difference is within 0.5 km, 0.5 is given if it is within 1 km, and 0.3 is given if it is within 10 km.

In S452, the situation similarity calculation unit 222 calculates the degree of coincidence in the direction from the departure place to the search point.
The situation similarity calculation unit 222 calculates the degree of coincidence in the direction from the departure place to the search point using, for example, the following equation.
In the following equation, Ds (u) indicates a 360 ° direction from the starting value of the user u to the search position.
Ds (u, v) indicates the difference in direction from the departure point of the users u and v to the search point.
R ₂ is a function that gives higher evaluation as the direction is closer.
For example, 1.0 is given if the difference is within 10 °, 0.5 is given if it is within 30 °, 0.3 is given if it is within 90 °.

In S453, the situation similarity calculation unit 222 calculates the degree of coincidence in the direction from the search point to the destination.
The situation similarity calculation unit 222 calculates the degree of coincidence of directions from the search point to the destination, for example, using the following formula.
In the following equation, Dt (u) indicates a 360 ° direction from the search point of the user u to the destination.
Dt (u, v) indicates a difference in direction from the search point of the user u and the user v to the destination.

In S454, the situation similarity calculation unit 222 adds the proximity of the search position, the degree of coincidence in the direction from the departure point to the search point, and the degree of coincidence in the direction from the search point to the destination, thereby adding the degree of situation similarity. And
In the following equation, φ, ψ, and ξ are adjustment parameters.

Note that, when the composite similarity V (u, v) is calculated in S460 of FIG. 6, if there is no relationship between the users u and v (for example, common between the users u and v in S423). When there is no destination group, the similarity propagation calculation unit 223 calculates the indirect similarity between the users u and v to obtain the combined similarity V (u, v).
That is, when the composite similarity V (u, v) between the user u who requested the recommendation request and a certain user v is less than a threshold value (including the case where V (u, v) = 0), Users m and n (corresponding to appropriate target selection history users) whose combined similarity V (v, m) and V (v, n) with the user v (corresponding to the inappropriate target selection history user) are equal to or greater than a threshold value. ) Exists, the combined similarity V (u, v) can be corrected.
For example, since there is no common destination between the users u and v, when V (u, v) = 0, the combined similarity V (u, m) and V (u, n) with the user u ) Are 0.5 and 0.1, respectively, and there are users m and n whose combined similarities V (v, m) and V (v, n) are 0.5 respectively. , 0.1.
In this case, V (u, v) = V (u, m) × V (m, v) + V (u, n) × V (n, v) = 0.25 + 0.01 = 0.26 Can do.
When there are many such indirect paths from user to user, the similarity is propagated by adding all of them to obtain the final similarity between users.

To obtain V (u, m), Su (u, m) and Ss (u, m) are required. To obtain V (u, n), Su (u, n) and Ss (U, n) is necessary, and in order to obtain V (v, m), Su (v, m) and Ss (v, m) are necessary, and in order to obtain V (v, n) Requires Su (v, n) and Ss (v, n).
As described above, Su (u, m), Su (u, n), Su (v, m), and Su (v, n) are calculated by the distance preference similarity calculation unit 221 according to the procedure shown in FIG. calculate.
Further, as described above, Ss (u, m), Ss (u, n), Ss (v, m), and Ss (v, n) are processed by the situation similarity calculation unit 222 according to the procedure shown in FIG. calculate.
Then, the preference / situation-considering content evaluation unit 224 calculates V (u, m) from Su (u, m) and Ss (u, m), and from Su (u, n) and Ss (u, n). V (u, n) is calculated, V (v, m) is calculated from Su (v, m) and Ss (v, m), and V (v, n) is calculated from Su (v, n) and Ss (v, n). v, n) is calculated.
Then, the similarity propagation calculation unit 223 calculates V (u, v) from V (u, m), V (u, n), V (v, m), and V (v, n).

In FIG. 6, before extracting the content information to be recommended in S430, the preference similarity by distance is calculated for all registered user combinations in S420.
Instead, after extracting the user who selected the recommendation target content information in S440 in the past, the preference similar by distance is extracted between the extracted user and the user who requested the recommendation request in the procedure shown in FIG. The degree may be calculated.

As described above, in the present embodiment, the activity base of the user is calculated, the proximity of the activity base between the users and the similarity of the behavior at the distance from the activity base are calculated, and the obtained similarity is calculated. The similarity between the recommendation target user and the user who has visited the candidate point is calculated by propagating the similarity through the neighboring users of the recommendation target user, and the candidate point is evaluated.
In this embodiment, the candidate point is evaluated based on the distance between the current search position and the destination and the degree of coincidence between the search position and the destination in the past search history of the user who has visited the candidate point. It has been corrected.
Further, in the present embodiment, from the starting point and destination of the current user and the position at the time of search, the starting point and destination of the past history of the user who has visited the candidate point, and the position at the time of search, The evaluation of candidate points is corrected based on the direction from the departure point to the search position and the degree of match of the route, the direction from the search position to the destination and the degree of match of the route, and the degree of coincidence of the distance from the search position to the destination. ing.
For this reason, it is possible to recommend a candidate point suitable for the user's preference and the user's current situation even in an area away from the user's active area.

As described above, in the present embodiment, the information recommendation apparatus including the following means has been described.
(1) Recommendation information management means for managing user search history and recommendation results;
(2) An information recommendation response program that evaluates a recommendation target content from a preference similarity according to distance between a recommendation target user and a user who has directly evaluated the recommendation target content;
(3) Distance-based thought similarity calculation means for calculating the similarity of actions at the distance from the activity base between users;
(4) Similarity propagation calculation means for calculating the similarity for all users from the recommendation target user by propagating through a neighboring user having abundant data matching the recommendation target user;
(5) A situation similarity calculation unit that evaluates the recommendation target content based on the degree of coincidence between the situation at the time of search and the situation at the time of other users' past searches;
(6) Using the content evaluation result obtained by the situation similarity evaluation means, the content evaluation result obtained by the preference similarity evaluation means is corrected, and the preference / situation consideration for generating content information to be provided to the recommendation target user Content evaluation means;

In addition, the information recommendation device according to the present embodiment is
It has been described that a preference similarity correction calculation unit for each distance that corrects an activity distance similarity based on the proximity of activity bases between users is provided.

In addition, the information recommendation device according to the present embodiment is
It has been described that the content evaluation correction means for correcting the content evaluation result based on the degree of coincidence between the search position at the time of search and the search position at the time of other users' past searches has been described.

In addition, the information recommendation device according to the present embodiment is
Content evaluation correction means for correcting the content evaluation result based on the degree of coincidence between the direction and route from the departure point to the search position at the time of search and the direction and route from the departure point to the search position at the time of past search by another user Explained that it is equipped with.

In addition, the information recommendation device according to the present embodiment is
Content evaluation correction means for correcting the content evaluation result based on the degree of coincidence between the direction and route from the search position to the destination at the time of search and the direction and route from the search position to the destination at the time of past search by other users Explained that it is equipped with.

Finally, a hardware configuration example of the information recommendation device 200 shown in the present embodiment will be described with reference to FIG.
The information recommendation device 200 is a computer, and each element of the information recommendation device 200 can be realized by a program.
As a hardware configuration of the information recommendation device 200, an arithmetic device 901, an external storage device 902, a main storage device 903, a communication device 904, and an input / output device 905 are connected to the bus.

The arithmetic device 901 is a CPU (Central Processing Unit) that executes a program.
The external storage device 902 is, for example, a ROM (Read Only Memory), a flash memory, or a hard disk device.
The main storage device 903 is a RAM (Random Access Memory).
The recommendation information storage unit 230 is realized by the external storage device 902 or the main storage device 903.
The communication device 904 corresponds to the physical layer of the communication unit 210.
The input / output device 905 is, for example, a mouse, a keyboard, a display device, or the like.

The program is normally stored in the external storage device 902, and is loaded into the main storage device 903 and sequentially read into the arithmetic device 901 and executed.
The program is a program that realizes a function described as “˜unit” (excluding the recommendation information storage unit 230, the same applies below) shown in FIG.
Further, an operating system (OS) is also stored in the external storage device 902. At least a part of the OS is loaded into the main storage device 903. ”Is executed.
In the description of the present embodiment, “determination of”, “determination of”, “extraction of”, “search of”, “calculation of”, “calculation of”, and “evaluation of” ”,“ Setting of ”,“ registration of ”,“ selection of ”,“ generation of ”,“ reception of ”,“ transmission of ”, etc. Data, signal values, and variable values are stored in the main storage device 903 as files.
Further, the encryption key / decryption key, random number value, and parameter may be stored in the main storage device 903 as a file.

Note that the configuration of FIG. 13 is merely an example of the hardware configuration of the information recommendation device 200, and the hardware configuration of the information recommendation device 200 is not limited to the configuration illustrated in FIG. Also good.
Also, the mobile terminal 100 shown in the present embodiment may have the hardware configuration shown in FIG. 13 or may have another hardware configuration.

  Further, the data processing method according to the present invention can be realized by the procedure shown in the present embodiment.

  DESCRIPTION OF SYMBOLS 100 Mobile terminal, 110 Communication part, 120 Information recommendation request process part, 130 Location information acquisition part, 140 Navigation process part, 200 Information recommendation apparatus, 210 Communication part, 220 Information recommendation response process part, 221 Calculation of preference similarity according to distance 222, situation similarity calculation unit, 223 similarity propagation calculation unit, 224 preference / situation consideration content evaluation unit, 225 recommendation information management unit, 226 information recommendation response unit, 230 recommendation information storage unit, 231 user information management table, 232 Content information management table, 233 search history management table, 234 recommendation result management table, 300 network.

Claims (12)

A search request indicating a start location, a destination, and a search request transmission location of a user moving from the start location to the destination is received from the user terminal device used by the user, and the destination and search indicated in the search request are received. An information processing device that searches content information related to at least one of request transmission locations and transmits a search response that notifies the searched content information to the user terminal device,
User base information indicating activity bases of a plurality of users, search request history information indicating the contents of a search request from a search request history user who has transmitted a search request among the plurality of users, A storage for storing selection history information indicating content information selected by a selection history user who has selected any content information from among the plurality of users notified by a search response. And
A content information search unit for searching for content information related to at least one of the destination and the search request transmission location indicated in the reception search request received from the user terminal device of any one of the plurality of users;
Based on the search request history information, the selection history information, and the received search request, the target selection history user who has selected the search content information searched by the content information search unit and the request source of the received search request A common destination with a requesting user, a distance between the common destination and the activity base of the requesting user indicated in the user base information, and the common destination and the user base information. A content information evaluation value calculation unit that calculates an evaluation value of the search content information using a distance difference between the target selection history user and the distance to the activity base shown;
An information processing apparatus comprising: a search response transmission unit configured to transmit a search response for notifying the search content information and the evaluation value calculated by the content information evaluation value calculation unit to the user terminal device. A search request indicating a start location, a destination, and a search request transmission location of a user moving from the start location to the destination is received from the user terminal device used by the user, and the destination and search indicated in the search request are received. An information processing device that searches content information related to at least one of request transmission locations and transmits a search response that notifies the searched content information to the user terminal device,
A user base information storage unit for storing user base information indicating activity bases of a plurality of users;
A search request history user that is a user who has transmitted a search request among the plurality of users, and a starting point, a destination, and a search request transmission location indicated in the search request transmitted by the search request history user A search request history information storage unit for storing search request history information shown in association with each other;
The selection history user who has selected any content information from the content information notified by the search response among the plurality of users and the content information selected by the selection history user are associated with each other. A selection history information storage unit for storing the displayed selection history information;
A search request receiving unit that receives a search request from a user terminal device of any one of the plurality of users;
A content information search unit that searches for content information related to at least one of a destination and a search request transmission location indicated in the received search request received by the search request receiving unit;
A selection history user who has selected the search content information searched by the content information search unit is extracted from the selection history information as a target selection history user, and when the target selection history user selects the search content information The target selection history user's departure location, destination, and search request transmission location are extracted from the search request history information, and the target selection history user departure location, destination, search request transmission location, and the received search are extracted. A situation similarity calculation unit that calculates the similarity between the origin and the destination of the request source user who is the request source of the reception search request indicated in the request and the search request transmission place;
A common destination between the request source user and the target selection history user is extracted from the search request history information, and the common destination and the activity base of the request source user indicated in the user base information The distance difference between the distance and the distance between the common destination and the activity base of the target selection history user indicated in the user base information is calculated, and the similarity based on the calculated distance difference is used as the distance similarity. A distance similarity calculation unit to calculate,
Using the situation similarity calculated by the situation similarity calculation unit and the distance similarity calculated by the distance similarity calculation unit, the similarity between the request source user and the target selection history user A content information evaluation value calculation unit that calculates a similarity between certain users and calculates an evaluation value of the search content information using the calculated similarity between users;
An information processing apparatus comprising: a search response transmission unit configured to transmit a search response for notifying the search content information and the evaluation value calculated by the content information evaluation value calculation unit to the user terminal device. The situation similarity calculation unit
When a plurality of pieces of content information are searched by the content information search unit, a target selection history user is extracted for each search content information, a situation similarity is calculated for each target selection history user,
The distance similarity calculation unit
Calculate distance similarity for each target selection history user,
The content information evaluation value calculation unit
For each target selection history user, the similarity between users with the request source user is calculated using the situation similarity and distance similarity of each target selection history user, and the similarity between users for each target selection history user The information processing apparatus according to claim 2, wherein an evaluation value is calculated for each search content information using a degree. The situation similarity calculation unit
When the same target selection history user is extracted with respect to a plurality of search content information, the target selection history user when the target selection history user selects the search content information is searched for each search content information. A departure place, a destination, and a search request transmission place are extracted from the search request history information, a situation similarity is calculated for each search content information, an average value of the situation similarities of the plurality of search content information is calculated,
The content information evaluation value calculation unit
For a target selection history user extracted in duplicate for a plurality of search content information, the average value of the situation similarity of the target selection history user and the distance similarity are used to The information processing apparatus according to claim 3, wherein similarity between users is calculated. The content information evaluation value calculation unit
When a plurality of target selection history users are extracted, for each target selection history user, calculate the similarity between users with the request source user,
When there is an inappropriate target selection history user whose user similarity with the request source user is less than a threshold among the plurality of target selection history users,
Calculating the similarity between users between the improper target selection history user and the appropriate target selection history user whose user similarity between the request source user is equal to or greater than a threshold;
Based on the similarity between users between the inappropriate target selection history user and the appropriate target selection history user and the similarity between users between the proper target selection history user and the requesting user, the inappropriateness Correct the similarity between users between the target selection history user and the request source user,
Using the inter-user similarity between the appropriate target selection history user and the requesting user and the correction value of the similarity between users between the inappropriate target selection history user and the requesting user, The information processing apparatus according to claim 2, wherein an evaluation value of search content information is calculated. The situation similarity calculation unit
The content information selected by the inappropriate target selection history user and the proper target selection history user is extracted from the selection history information, and the inappropriate target selection history user selects the content information. The starting point, the destination, and the search request transmission place of the inappropriate target selection history user are extracted from the search request history information, and the proper target selection history user when the appropriate target selection history user selects the content information. A departure place, a destination, and a search request transmission place are extracted from the search request history information, and the extracted inappropriate target selection history user's departure place, destination, search request transmission place, and the extracted proper target selection history are extracted. Calculate the similarity between the user ’s departure point, destination, and search request sending location as the situation similarity,
The distance similarity calculation unit
A common destination between the inappropriate target selection history user and the proper target selection history user is extracted from the search request history information, and the inappropriate target selection indicated in the common destination and the user base information A distance difference between the history user's activity base and the distance between the common destination and the appropriate target selection history user's activity base shown in the user base information is calculated, and the similarity based on the distance difference Calculating the degree as a distance similarity,
The content information evaluation value calculation unit
Using the situation similarity calculated by the situation similarity calculation unit and the distance similarity calculated by the distance similarity calculation unit, between the inappropriate target selection history user and the appropriate target selection history user The information processing apparatus according to claim 5, wherein the similarity between users is calculated. The situation similarity calculation unit
The distance between the search request transmission location of the target selection history user and the search request transmission location of the request source user, the direction from the departure location of the target selection history user to the search request transmission location, and the departure location of the request source user Analyzing the direction to the search request transmission location, the direction from the search request transmission location to the destination of the target selection history user, and the direction from the search request transmission location to the destination of the requesting user, and similar to the situation The information processing apparatus according to claim 2, wherein a degree is calculated. The distance similarity calculation unit
Before the target selection history user is extracted by the situation similarity calculation unit,
A search request history user pair is generated by combining the search request history users, and a common destination among the search request history users is extracted from the search request history information for each search request history user pair, and the common purpose is extracted. Calculate the distance difference in the distance between the ground and the activity base of each search request history user shown in the user base information, calculate the distance similarity between the search request history users of each pair based on the calculated distance difference,
When the target selection history user is extracted by the situation similarity calculation unit,
The information processing apparatus according to claim 2, wherein a distance similarity with respect to a pair of the request source user and the target selection history user is selected. A search request indicating a start location, a destination, and a search request transmission location of a user moving from the start location to the destination is received from the user terminal device used by the user, and the destination and search indicated in the search request are received. An information processing method performed by a computer that searches content information related to at least one of request transmission locations and transmits a search response that notifies the searched content information to the user terminal device,
A search request in which the computer shows user base information indicating activity bases of a plurality of users, and contents of a search request from a search request history user who has transmitted a search request among the plurality of users. History information and selection history information indicating content information selected by a selection history user who has selected any content information from among the plurality of users notified by a search response. Reading information from the storage area;
A content information search in which the computer searches for content information related to at least one of a destination and a search request transmission location indicated in a reception search request received from a user terminal device of any one of the plurality of users. Steps,
Target selection history user who has selected the search content information searched by the content information search step based on the search request history information, the selection history information, and the reception search request, and the reception search request A common destination among the requesting users who are requesting users, a distance between the common destination and the activity base of the requesting user indicated in the user base information, the common destination and the A content information evaluation value calculating step of calculating an evaluation value of the search content information using a distance difference between the target selection history user and the activity base indicated in the user base information;
The computer has a search response transmission step of transmitting a search response for notifying the search content information and the evaluation value calculated by the content information evaluation value calculation step to the user terminal device. Information processing method. A search request indicating a start location, a destination, and a search request transmission location of a user moving from the start location to the destination is received from the user terminal device used by the user, and the destination and search indicated in the search request are received. An information processing method performed by a computer that searches content information related to at least one of request transmission locations and transmits a search response that notifies the searched content information to the user terminal device,
A user base information reading step in which the computer reads user base information indicating activity bases of a plurality of users from a storage area;
A search request history user that is a user who has transmitted a search request among the plurality of users, and a departure point, a destination, and a search request indicated in the search request transmitted by the search request history user A search request history information reading step for reading search request history information indicated in association with a transmission location from a storage area;
A selection history user that is a user who has selected any content information from among the plurality of users notified by a search response among the plurality of users, and content information selected by the selection history user; A selection history information reading step for reading selection history information indicated in association from the storage area;
A search request receiving step in which the computer receives a search request from a user terminal device of any one of the plurality of users;
A content information search step in which the computer searches for content information related to at least one of a destination and a search request transmission location indicated in the received search request received by the search request receiving step;
The computer extracts a selection history user who has selected the search content information searched in the content information search step as a target selection history user from the selection history information, and the target selection history user extracts the search content information. The target selection history user's departure point, destination, and search request transmission point when selected are extracted from the search request history information, and the target selection history user's departure point, destination, and search request transmission point are extracted. A situation similarity calculation step of calculating a similarity between a departure place and a destination of a request source user who is a request source of the reception search request indicated in the reception search request and a search request transmission place as a situation similarity. ,
The computer extracts a common destination between the request source user and the target selection history user from the search request history information, and the request source user indicated by the common destination and the user base information A distance difference between the distance to the activity base and the distance between the common destination and the distance between the target selection history user activity base indicated in the user base information is calculated, and the similarity based on the calculated distance difference is calculated. A distance similarity calculating step for calculating the distance similarity;
The computer uses the situation similarity calculated in the situation similarity calculation step and the distance similarity calculated in the distance similarity calculation step, between the request source user and the target selection history user. A content information evaluation value calculating step of calculating an evaluation value of the search content information using the calculated similarity between users,
The computer has a search response transmission step of transmitting a search response for notifying the search content information and the evaluation value calculated by the content information evaluation value calculation step to the user terminal device. Information processing method.   A program for causing a computer to function as the information processing apparatus according to claim 1.   A starting point, a destination, and a search request transmission place of a user who moves from the departure place to the destination are indicated, a search request for requesting search for content information is transmitted to the information processing apparatus, and search content information for the search request , Receiving an evaluation value of the search content information based on the departure point, destination, and search request transmission location of another user who has selected the search content information from the information processing device, and A user terminal device comprising an information recommendation request processing unit for displaying the search content information based on the information recommendation request processing unit.

Images