JP2014215670A - Reliability calculation device - Google Patents

Abstract

A reliability calculation device for calculating a reliability suitable for determination of a partner of a vehicle is provided. An address acquisition unit 11 acquires address information of other users registered in each user's terminal. The user map generation unit 12 generates a user map representing a human relationship by counting the acquired addresses, setting each user as a node, and providing an edge between users registered in the address. The reliability calculation unit 13 calculates the reliability of other users in each user as a higher value as there are more connections on the user map. Furthermore, matching between the users based on the reliability is also possible by the matching unit 14, and it is preferable to target users connected in two steps in the vehicle riding. [Selection] Figure 3

Description

  The present invention relates to a reliability calculation device, in particular, it is possible to automatically calculate the reliability of an opponent sharing a predetermined event, such as a partner of a vehicle, and further using the reliability, The present invention relates to a reliability calculation apparatus capable of presenting matching partner candidates.

  In vehicle sharing (ride sharing / car sharing), it is necessary to determine the members to use, conditions for use, etc.

  Regarding the vehicle riding, Patent Literature 1 discloses a riding mediating apparatus. Sharing information such as departure point / arrival point information, departure time / arrival time, etc., and exchanging information by e-mail between members realizes sharing.

JP 2003-281237 A

  In such a case, security problems between users are extremely large. Many conventional riders solve problems such as security, focusing on the convenience of the person driving the car. For example, in the following Patent Document 2 (authentication system in a car sharing system), an IC driver's license is used to authenticate a user.

  [Patent Document 2] Japanese Patent Application Laid-Open No. 2012-247818

  However, there are cases where you want to use a car without driving. In such a case, the technique of Patent Document 2 aims to improve the convenience of the person who drives the car, and is not considered to improve the reliability of the person who gets on the vehicle. Therefore, the reliability of the person who gets on the vehicle cannot always be ensured, and in some cases, it is necessary to ride with a person with low reliability, which is a barrier to use for general users.

  In addition, when determining the partner or group to share a predetermined event from a large number of unspecified members, such as when a community is formed between participants of a so-called town con, not only for riding, reliability is also increased. It is desirable to be able to secure and determine opponents and groups. At this time, each member is a user who uses a predetermined system, and it is desirable that the reliability can be automatically determined and determined by the system.

  In view of the problems as described above, a first object of the present invention is to provide a reliability calculation device capable of calculating the reliability of the other party.

  The second object of the present invention is to realize matching between users based on the calculated reliability.

  In order to achieve the above object, the present invention is a reliability calculation device for calculating the reliability of other users in each user via each user's terminal, and the terminals of other users registered from each terminal An address acquisition unit that acquires address information, and a user map generation unit that generates a user map as a graph in which each user is a node and an edge is provided between users with address registration using the acquired address, A first feature is that a reliability calculation unit that calculates the reliability of other users in each user based on a path from the user to the other user using the generated user map is provided.

  In the present invention, the reliability calculation device further presents matching candidate users to the requesting user based on the calculated reliability to the user terminal requesting matching, A second feature is that a matching unit that mediates matching between terminals is provided.

  According to the first feature, a user map is automatically generated reflecting the registered relationship between users at the address collected from each terminal, that is, the human relationship between users, and the path between the users on the user map is Calculate reliability. Therefore, the reliability reflecting the human relationship can be automatically calculated, and the first object is achieved.

  According to the second feature, since matching between users is realized by the calculated reliability, the second object is achieved.

It is a block diagram of the reliability calculation system which concerns on one Embodiment. It is a figure for demonstrating the outline | summary of the reliability calculation and matching in this invention. It is a functional block diagram of the reliability calculation apparatus which concerns on one Embodiment. It is an example of a user map. It is an example of the address information of each terminal used for the user map generation of FIG. It is an example of a user map for showing an example of matching. It is an example of a series of screens that a determination unit presents to each user during matching in the first embodiment. It is an example of a series of screens that a determination unit presents to each user during matching in the second embodiment.

  FIG. 1 is a configuration diagram of a reliability calculation system according to an embodiment. The reliability calculation system 100 includes a reliability calculation device 10 and terminals 1A, 1B, 1C,... Owned by each user A, B, C,. Each terminal 1A and the like can typically be configured by a portable information terminal such as a smartphone. The reliability calculation device 10 may be configured as a server, or may be configured in any one of the terminals 1A and the like. The outline of the present invention according to this configuration is as follows.

  Each user A or the like transmits the address information of other users registered from his / her terminal 1A or the like to the reliability calculation apparatus 10. The reliability calculation device 10 aggregates the address information transmitted from each user A and calculates the reliability of other users in each user A and the like as the first process. Further, as a second process, the reliability calculation device 10 mediates matching between users based on the calculated reliability, requests from each user A and the like, and pre-registration information.

  FIG. 2 is a diagram for explaining the reliability calculated in the present invention and the outline of matching based on the reliability. The reliability calculation device 10 aggregates the address information transmitted from each of the users A and constructs human relationship information (user map described later) as conceptually shown in FIG. The human relationship information is information that identifies a user who is in a friendship relationship (hereinafter referred to as “connection”).

  In the present invention, the reliability is calculated on the assumption that the user having more connections has higher reliability. When matching is performed based on reliability, matching is performed in descending order of reliability for users who are connected in two steps. The connection in two steps is a relationship of “friends of friends”, and is a relationship in which there is no direct connection (connection in one step) but there is a common friend. In addition, the matching conditions for sharing the event are performed at the time of the matching, and matching is performed in descending order of reliability among users who meet the conditions.

  FIG. 2 shows the human relationship information when the user a is viewed as a center by drawing a double-headed arrow between users who have a connection in one step. As shown in the figure, three users who have a direct connection (connection in one step) with the user a are b, c, and d. Each of the users b, c, and d has a connection in one step from itself, and when viewed from the user a, the users b1, b2, b3 and c1, c2, c3 have a connection in two steps. , D1, d2, and d3.

  Therefore, when matching with the user a, the nine users b1, b2, b3, c1, c2, c3 and d1, d2, d3 are candidates for matching, and the nine users Among them, those that meet the conditions for sharing the event are selected as matching targets in descending order of reliability.

  In the present invention, in particular, the following effects can be achieved by setting a user having a connection in two steps as a matching candidate. First, a highly reliable candidate is automatically selected based on the presence of a common friend among an unspecified number of users. Secondly, since the number of candidates is approximately the square of the number of users who are directly connected, and there are a large number, there are many users who meet the matching conditions.

  In addition, from the viewpoint of securing a larger number of users that match the matching conditions by relaxing the reliability condition, users having connections in three or more steps may be added to the matching candidates. Even in such a case, since the reliability is calculated for each candidate, the candidates can be presented in descending order of reliability.

  In addition, a user who has a connection in one step can be directly exchanged between the users, and is not included in the matching candidates. Since there are many users who have connections in two or more steps, there is a large number of users, but this is significant as a matching candidate according to the present invention.

  FIG. 3 is a functional block diagram of the reliability calculation apparatus 10 according to an embodiment. The reliability calculation device 10 includes an address acquisition unit 11, a user map generation unit 12, a reliability calculation unit 13, and a matching unit 14. Here, the matching unit 14 further includes a condition acquisition unit 141, a candidate presentation unit 142, and a determination unit 143.

  3 is drawn including the interaction with each user terminal 1A, 1B, 1C,... As shown in FIG. For convenience of explanation, in FIG. 3 and the following explanation, the user who requests the matching is the owned user A of the terminal 1A and the other users B, C ... (terminals 1B, 1C, ...). Is the side accepting and accepting the matching request.

  The address acquisition unit 11 acquires registered addresses from the user terminals 1A, 1B, 1C,. The address is a contact information of another user, and the following can be used. If each terminal is a mobile phone, the phone number may be used. In addition, an e-mail address in each terminal may be used. Alternatively, an ID of an SNS (social networking service) used by the user at each terminal may be used. When acquiring the SNS ID, it may be acquired from the SNS server or the like instead of from each terminal.

  Further, the name of each user and the terminal ID may be acquired in association with the address. The name or the like can be used as information indicating a user when mediation between users is performed by the matching unit 14 described later.

  The user map generation unit 12 uses the registered address acquired by the address acquisition unit 11 to determine that there is a connection between users, and generates a user map that represents the connection between users. To do. The user map is generated as a graph in which each user is a node and an edge is provided between users who have connections due to the presence of address registration.

  When determining the presence of a connection between users from address registration, it may be determined that there is a connection when there is mutual address registration between both users, and at least one of them may be determined. If there is an address registration, it may be determined that there is a connection.

That is, the connection between the users x and y may be determined to exist when both of the following conditions (1) and (2) are satisfied, and either of the conditions (1) or (2) It may be determined that a connection exists when at least one of the conditions is satisfied. The “address book” refers to a means (memory area or the like) for storing an address registered in each terminal in the terminal.
(1) y is registered in the address book of x
(2) x is registered in the address book of y

  In addition, when determining the connection in at least one of the conditions (1) or (2) (unidirectional), when there is registration of both (bidirectional), the reliability between users in the reliability calculation unit 13 described later May be calculated higher than in the case of a single direction. Details of the calculation will be described later.

  In the user map generation unit 12, all of the registered addresses acquired by the address acquisition unit 11 may be aggregated to generate a user map in which the connections of all users are expressed collectively. Alternatively, the user map generating unit 12 refers to the registered address acquired by the address acquiring unit 11 after receiving a matching request from the terminal 1A of the specific user A (as indicated by a dotted arrow in FIG. 3). Thus, the user map may be generated within a predetermined range centered on the specific user A.

  FIG. 4 is an example of a user map. FIG. 5 is an example of address information registered in each terminal used to generate the user map of FIG. Here, as a partial example, there are shown users B to I around the user A who are connected to each other.

  The reliability calculation unit 13 uses the user map generated by the user map generation unit 12 to calculate the reliability of other users in each user. In particular, when a request for matching is received from the terminal 1A of the specific user A, the reliability is calculated for each of the other users viewed from the specific user 1A. The calculation result may be notified to the terminal of the user who is the starting point of the reliability calculation (or the user A who requested the matching request if there is a matching request).

  At the time of the calculation, the reliability calculation unit 13 refers to the user map, groups other users for reliability calculation according to the number of connection steps from the user that is the starting point of the reliability calculation, Every time, the reliability of other users is calculated. The connection step number is a minimum value of the number of steps required to move between nodes representing the user in the user map.

In the example of FIGS. 4 and 5, if the reliability of each of the other users B to I viewed from the user A is calculated, it is distinguished from (0), (1), (2) in FIG. In addition, the reliability is calculated separately for each group as follows.
User groups connected in one step: B, H, G, I, F
User groups connected in two steps: C, D, E, ...
(Similarly, it may be calculated for each group of users connected in three steps or more.)

  The reliability for users connected in one step is calculated according to the following policy.

  As a first policy, the more the paths other than the minimum one step between users, the more common friends, the more intimate relationship between the users can be determined, so the reliability value is calculated higher . Here, for paths other than one step, paths up to a predetermined number n (n ≧ 2) steps are considered, and the degree of addition to the reliability is increased as the path has a smaller step number i (2 ≦ i ≦ n). .

  For example, in the example of FIG. 4, for user B, which is connected in one step as viewed from user A, there is a two-step path for AGB and AHB in addition to a one-step path for AB. Calculate the reliability of user B viewed from the above.

  As the second policy, the more frequently the number of common friends having a path up to the predetermined number n (n ≧ 2) steps in the first policy is updated and increased, the more the common friends interact with each other. Therefore, the reliability between the users is calculated to be higher.

According to the first and second policies described above, specifically, the reliability tr1 for a user connected in one step can be calculated as follows, for example. As a value used for the reliability tr1, the number of paths that can be reached between the users with the minimum number of times i (i ≧ 2) is s _i . For example, in the case of calculating the reliability of the user B viewed from the user A in the example of FIG. 4, since AGB and AHB exist, s ₂ = 2 between the users A and B.

In addition, as a value used for the reliability tr1, when a friend who finally reaches the i-th time is added among the users (the connection book of the i-th time is newly generated by updating any user's address book) Let t _{i be} the elapsed time from the most recent time) to the present time. For example, when calculating the reliability of the user B as viewed from the user A in the example of FIG. 4, the date and time when the connection relationship between the AGB and AHB has occurred is T (G) and T (H), and the current date and time is T _{Assuming now} , t ₂ is the smaller of {T _now −T (G)} and {T _now −T (H)}.

The reliability tr1 can be calculated by, for example, the following (Equation 1) using the number of passes s _i and the elapsed time t _i and the predetermined positive monotone increasing function f and positive monotonic decreasing function g. As described above, the constant n (n ≧ 2) is the upper limit of the step number i of the connection to be considered.

That is, in (Expression 1), the function f is a term of the first policy, and the reliability increases as the number of paths s _i increases. Further, the function g is a term of the second policy, and the reliability is increased as the elapsed time t _i from the latest update time is shorter.

Note that the elapsed time t _i is acquired from the date and time information when the address acquisition unit 11 acquires the address information including information on the date and time when the address of the other terminal is registered in each terminal. be able to. Alternatively, the address information acquisition by the address acquisition unit 11 and the user map generation by the user map generation unit 12 are periodically performed, the user map is updated periodically and the history is recorded, thereby calculating the current reliability. You may make it obtain | require with reference to the said log | history the latest past when the part relevant to the object user became the present state.

As for the second policy, the one obtained by replacing g (t _i ) in (Expression 1) with the following term may be used. That is, based on the above history, the frequency at which the i connection relationships are added and updated is quantified by a predetermined method to be α _i, and by A (α _i ) using a predetermined positive monotonically increasing function A G (t _i ) in (Equation 1) may be replaced.

In addition, the reliability tr1 may be calculated from only the first policy without using the second policy. In this case, g (t _i ) = 1 may be substituted in (Expression 1). For example, in the example of FIG. 4, when the reliability is calculated for only the first policy for the users B, H, G, I, and F connected from the user A in one step, the magnitude relationship is as follows.
B> H = G> I = F

  That is, when viewed from A, B has two relations connected in two steps, AHB and AGB, and H and G have one relation connected in two steps, ABH and ABG, respectively, and I and F Since there is no relationship that connects in two steps, there is a size relationship as described above. In this example, the upper limit step number n = 2.

  The reliability for a user connected in two steps can be calculated by applying a third policy in addition to the first and second policies applied in the user connected in one step.

  In the third policy, taking into account the reliability of a common friend that is routed between users connected in two steps (the reliability calculated for a user connected in one step), the higher the reliability, the more the two steps The reliability between connected users is also high.

  A value obtained by adding the reliability tr1 [j] obtained by (Equation 1) for the common friend j that is connected when connecting in the two steps is added to all j is TR. Here, when j = 1, 2,... And there is a connection relationship by a plurality of two steps, it is indicated by j that there is one or more common friends. By using a predetermined positive monotone increasing function h for reflecting the third policy, specifically, for example, the reliability tr2 between users connected in two steps can be calculated by the following (Equation 2). .

Alternatively, as in the following (Equation 2-b) and (Equation 2-c), when obtaining the number of paths s _i , the value of s _{i is} set in consideration of the reliability of the user passing through each path. By modifying to ', the reliability tr2 reflecting the third policy may be obtained. Note that tr _{i, j} in (Equation 2-b) is the reliability for each via user j in each path in the number of paths s _i , and i ≧ 3 is calculated in advance by one of the methods described below You just have to.

Note that the application of the second policy and / or the third policy may be omitted, and the reliability tr2 between users connected in steps may be calculated. In this case, replacement such as g (t _i ) = 1 or h (TR) = 1 may be performed in (Expression 2) or the like.

  When obtaining the reliability trk between users connected by k steps of 3 steps or more, it may be calculated from the following (Equation 20) in the same manner as (Equation 2) between users connected by 2 steps. Here, the value TR to be added as the reliability of the common friend that passes through is obtained using the value calculated in k−1 steps or less, and the sum (Σ) of the product terms of the functions f and g is ( What is necessary is just to obtain | require in the range from i = k to i = n as described in Formula 20). Thus, the reliability between users connected in 3 steps or more can be calculated in order of 3 steps, 4 steps,... (N−1) steps, and n steps. Alternatively, it may be calculated by the method of (Expression 2-b) and (Expression 2-c) instead of (Expression 20).

  If the third policy is not applied, the user map (and the update date / time information) is not referred to when the reliability is calculated between the users connected in k steps without referring to the reliability TR of (k−1) steps or less. ) Only. In this case, h (TR) = 1 may be substituted in (Equation 20).

  When calculating the reliability between users at each number of steps as described above, the fourth policy is to consider the frequency of communication between users and increase the reliability so that the higher the communication frequency, the higher the value. You may make it calculate.

  Regarding the communication frequency, when the address acquisition unit 11 acquires the address of another user registered by each user, as additional information, as a communication history between the addresses of each user, for example, the number of communication over a predetermined period, etc. Get it. Thus, the communication frequency between users connected in one step is obtained as a numerical value.

  In order to reflect the communication frequency obtained as the numerical value in the calculation of the reliability according to the fourth policy, for example, the following method can be used. Only one or both of the following first and second methods may be used.

  As a first method, instead of (Equation 1), for example, a modification as shown in (Equation 3) below can be used. Here, u is a numerical value of the communication frequency between users whose reliability is to be calculated in (Expression 3), and F is a predetermined monotonically increasing function.

As a second method, the number of paths s _i that can be reached with the minimum number of times i (i ≧ 2) in (Expression 1) or (Expression 3), (Expression 2), or (Expression 20) As in 4), a value corrected based on the sum U _i of the communication frequency values u between the two users in the s _i paths can be obtained. Here, G is a predetermined positive monotonously increasing function and takes a value larger than 1. The corrected value is not necessarily an integer. Note that the following correction method of (Equation 4) is an example, and the communication frequency may be reflected in the value of s _i by other methods. The value of s _i may be corrected by using (Formula 2-b) and (Formula 2-c) together.

Alternatively, instead of using the number of paths s _i modified as in (Equation 4), G (U _i ) itself in (Equation 4) may be added to the terms of the function f and g. . For example, in the case of (Equation 2), the following (Equation 5) may be modified.

As a fifth policy, the contribution to reliability by the term of the number of paths s _{i in} the above is doubled when the above-described address registration is not unidirectional but bidirectional in the description of the user map generation unit 12, for example. For example, the reliability may be calculated higher than when only one direction is registered. In this case, using a method similar to (Equation 4), the value increases as the ratio increases, based on the ratio of two-way registration among the two users in the s _i paths. Thus, the value of s _i may be corrected.

  The reliability calculation unit 13 has been described above. Hereinafter, returning to FIG. 3, each part will be described.

  The matching unit 14 provides matching to the terminal 1A of the user A requesting the matching by presenting a matching partner based on the reliability of other users viewed from the user A calculated by the reliability calculation unit 13. Intermediary. For the mediation, each unit 141, 142, 143 performs the following processing.

  The condition acquisition unit 141 acquires conditions imposed on the other party when sharing events between users from each terminal. The condition may be specified by dividing a predetermined item and allowing the user to select whether to request an event sharing partner for each item. Such predetermined items may be selected from each user in advance before accepting a matching request. Further, there may be an item based on predetermined data acquired from a terminal or the like at the time of request.

  For example, if the event is a vehicle ride, the following items are specified as the predetermined items to be selected in addition to the items that specify whether each user provides the vehicle or the vehicle. May be provided. As conditions of conditions that the vehicle providing user imposes on the user of the partner, the date and region where the vehicle can be provided and the attributes of the partner user (gender, age, address, presence or absence of a driver's license, etc.) are set. Good. On the contrary, as a condition from the side of the user who receives the provision of the vehicle, the date / time and the region where the provision of the vehicle is desired (the desire to participate) and the attributes of the providing user may be set.

  Here, the area may be set separately for the departure place and the arrival place. You may set the attribute of oneself as a standpoint of a providing user and a providing partner user so that it can utilize when referring to and judging a providing user's attribute and a providing user's attribute from others. In addition, various items disclosed in Patent Document 1 may be set.

  In addition, about each item, when the said item is imposed as conditions, you may be comprised so that it can specify by selecting an applicable item from a predetermined menu. For example, the area may be specified by selecting from menu items such as Kanto and Kansai. When the item is not requested, a menu item indicating that may be selected.

  Further, in the case of a vehicle ride, the position information of each terminal may be used as predetermined data acquired from the terminal by the condition acquisition unit 141 at the time of request. In addition, each terminal acquires position information by GPS (Global Positioning System) or the like. If the user is a vehicle providing user, the position information may be acquired by a device (such as a car navigation system) provided in the vehicle to be provided.

  The location information indicates that a user who requests to ride a vehicle is a user who is present at a nearby location where the vehicle can be provided immediately after the request (by accident, driving to a common destination nearby) It is preferable to use this when searching for drivers and the like and agreeing between the users, and then quickly meeting with each other. In the case where a request is made as a reservation for carrying out vehicle riding in the future (for example, after one week), the position information is not necessarily required.

  In the case of carrying out immediate boarding as described above, the request side inputs the fact to the condition acquisition unit 141, and the providing side makes such a request for a predetermined period (if the conditions of the location information are met). Whether or not the request can be handled may be set in the condition acquisition unit 141. With this setting, the candidate presenting unit 142 described below can automatically select a user who matches a condition.

  The candidate presenting unit 142 sequentially receives the user A acquired by the condition acquisition unit 141 from the user with the higher reliability viewed from the user A calculated by the reliability calculation unit 13 with respect to the user A who makes the matching request. The user A terminal 1A is notified as to whether or not the conditions of both the high-reliability user and the high-reliability user are satisfied, and a predetermined number of matching conditions are listed. The confirmation of the match can be automatically confirmed by checking whether all the predetermined items set by both users match. In the notification of the list, the reliability value or its rank / rank, and the items acquired by the condition acquisition unit 141 in relation to matching may be notified together.

  When using the above-described position information by GPS or the like, the candidate presenting unit 142 may determine that the match is satisfied when the distance between users determined by the position information is within a predetermined value. The predetermined value may also be set in advance by each user with respect to the condition acquisition unit 141. In this case, when different distances are set for two users, the smaller value may be used for the match determination. Alternatively, a map in which a predetermined area is divided may be prepared, and when both users are in the same area in the position information, it may be determined as a match.

  As described above, since the one-step connection is a direct friend, it is preferable to select the matching candidates notified to the terminal 1A from the users connected in two steps in descending order of reliability. When a predetermined number of users who meet the conditions does not reach among the users connected in two steps, the users connected in three steps may be selected in the order of high reliability. In the case where the number is insufficient, the user may be selected from four or more steps. The same applies thereafter.

  For example, when user A makes a matching request in the example of FIG. 4, a user who satisfies the condition among the friends C, D, and E connected in two steps has a reliability of terminal A of user A. The list is notified in descending order. At this time, even if a list is presented together with information such as a candidate user's profile (a common friend is B, current position information, etc.) Good. It is not always necessary to notify the reliability rank. Instead of the rank, a rank such as “high” or “normal” may be notified, or a reliability value may be notified.

  Note that the notification of the candidate list by the candidate presenting unit 142 can take various forms. For example, the terminal 1A may be notified by sending it in the form of e-mail or the like, or on the website provided by the reliability calculation device 10, by the web page generated for the matching of the terminal 1A You may make it notify.

  The confirmation unit 143 confirms a final agreement regarding event sharing between the terminal 1A of the user A that has received the notification of the candidate list and the user terminal selected from the user A as a partner to be matched. The matching is confirmed by mediating through the terminals of both users. The confirmation flow is as shown in the following steps (S1) and (S2).

  (S1) The requesting user A selects a partner x for which matching is desired from the list of candidate users, and notifies the determination unit 143 from the terminal 1A.

  (S2) The confirmation unit 143 establishes communication between the terminal 1x of the selected user x and the terminal 1A of the requesting user A, and confirms whether or not both agree on matching. If the agreement is made, the determination unit 143 receives the fact and the flow ends. If the agreement is not reached, after the confirmation unit 143 receives the notification, the user A is notified that another partner is selected as the partner x, and the process returns to step (S1) and repeats.

  The supplementary items of the flow are as follows.

  When establishing communication in step (S2), both users A and x are notified of the addresses of the other parties according to the selection result x of the requesting user A (same type as acquired by the address acquisition unit 11), and The agreement may be confirmed by corresponding communication means (if the address is a telephone number, a telephone). Further, the address may not be notified, and mediation of agreement confirmation may be performed via the determination unit 143 (by a dedicated communication means such as a website provided by the reliability calculation device 10).

  In the case of vehicle riding, it is also possible to decide how to meet each other between the users A and x at the time of confirmation of the step (S2).

  In the case of vehicle riding, the position information obtained by the GPS or the like acquired by the condition acquisition unit 141 is independent of whether or not the candidate presentation unit 142 uses the position information to determine whether the condition is met, as follows. May be used.

  First, when confirming the agreement in step (S2) through the intermediation of the determination unit 143, as the information for the requested user x side to determine whether to agree, the requested terminal 1x The location information of the requesting terminal 1A may be provided. Second, in the requesting terminal 1A, when the candidate presenting unit 142 creates a candidate list to be used as reference information when selecting from the candidate list in step (S1), the candidate list includes candidates for the candidate list. You may make it include a user's positional information.

  As described above, as an example of the first embodiment by the matching unit 14, an example has been described in which mediation is performed between the user A who requests matching and the user who becomes a matching candidate (a user separated by two steps or more). Next, as a second embodiment by the matching unit 14, an example will be described in which matching is performed by additionally using a recommendation from a user connected in one step which is a direct friend of the requesting user A.

  In addition, in order to make the recommendation possible, in the second embodiment, a user as a matching candidate is configured only by users separated by two steps, and does not include users separated by three steps or more.

  In the second embodiment, the condition acquisition unit 141 performs the same processing as that of the first embodiment, and the candidate presentation unit 142 and the determination unit 143 perform processing different from that of the first embodiment through a direct friend. Since this process includes repetition until a matching partner is determined, the following steps (S10) to (S50) will be described separately.

  (S10) The candidate presentation unit 142 obtains a predetermined number of candidate users as a list in the same manner as in the first embodiment, and unlike the first embodiment, instead of the candidate users described in the list, the candidate presentation unit 142 is connected to the candidate users. Is notified to the terminal 1A that requests matching.

  (S20) In the terminal 1A, the requesting user A selects the friend X that he / she wants to request the recommendation of the matching partner from the direct friend list (recommendation request list), and notifies the determination unit 143 of it.

  (S30) The determination unit 143 extracts a user connected to the user A via the friend X from the list of candidate users obtained by the candidate presentation unit 142 as a recommended candidate list, and the terminal 1X of the friend X The recommendation candidate list is transmitted to make a recommendation.

  (S40) The friend X selects the matching partner Y recommended for the friend A from the recommendation candidate list, and notifies the determination unit 143 of it. (Note that the recommendation candidate list includes a direct friend of the user X, and the recommendation requester A is also a direct friend of the user X. It is easy to decide.)

  (S50) The confirmation unit 143 mediates agreement confirmation between the terminal 1A of the user A and the terminal 1Y of the recommended matching partner Y in the same manner as in step (S2) of the first embodiment. If the agreement is reached, the determination unit 143 receives the message and the flow ends. If the agreement is not reached, the same process is repeated until the agreement is reached according to the case classification (C1) or (C2) below as a case classification different from step (S2).

  (C1) If other users remain in the recommendation candidate list of friend X who requested the recommendation, the process returns to the above step (S40), and the determination unit 143 prompts the friend X to recommend again from the remaining users. Thereafter, the same processing is performed.

  (C2) If there are no other users remaining in the recommendation candidate list of friend X who requested the recommendation, the process returns to step (S20), and confirming unit 143 sends a list of direct friends to matching request user A ( After selecting a friend other than the common friend X from the recommendation request list), the same processing is performed thereafter.

  In the above steps, the friend “X” and the partner “Y” are “variables” for storing the user names remaining at the time of each step. Accordingly, the “X” and “Y” are replaced with the corresponding remaining user names when repeated processing is performed in accordance with the convention for describing program processing. Similarly, in (C2) or the like, the recommended partner is selected from all the friends who have already been requested.

  In addition, when extracting the recommendation candidate list in step (S30), the reliability is extracted in order of the reliability described in the list of candidate users and used for the determination in step (S40) for friend X. You may make it convey the information of order.

  FIGS. 6, 7, and 8 are diagrams for illustrating matching examples of the first embodiment and the second embodiment described above. FIG. 6 is a user map used in this example, and G friends J, K, and L are further added to the user map of FIG. 7 and 8 are examples of screens presented to the corresponding users by the determination unit 143 in the first embodiment and the second embodiment, respectively, when the user A requests matching. In the example of FIGS. 7 and 8, for the user A, in both of the embodiments, the conditions of the six persons C, D, E, J, K, and L that are connected in two steps are matched with each other. Assume that four users J and K are candidate users.

  FIG. 7 is an example of a series of screens presented to each user in the first embodiment. In the screen G10, candidate users C, D, J, and K are presented to the matching request user A. In the screen G11, when the user A selects the candidate user C, the user G is requested to confirm whether or not to permit the user C. The screen G12 informs the user A that the license has been granted when the user C grants the license. The screen G13 prompts the user A to select from the remaining candidate users D, J, and K excluding C when the user C does not permit.

  FIG. 8 is an example of a screen presented to each user in the second embodiment. In the screen G20, friends B and G who can recommend candidate users to the matching request user A are presented. (Here, among the candidate users C, D, J, K, the friend of C, D is B, and the friend of J, K is G.) On the screen G21, the user A is recommended to the friend B Is requested from the candidate users C and D (that is, friends of B) to the user B.

  In the screen G22, when the user B selects the candidate user C as a recommendation target, the user G is requested to confirm whether or not the matching with the user A is permitted by the recommendation of the user B. The screen G23 informs the user A that the user C has given permission. The screen G24 prompts a recommendation from the candidate user D remaining in the user B when the user C does not permit.

  If the user B recommends the candidate user D after the screen G24 and the user D does not permit, all the users C and D in the recommendation candidate list of the user B are disapproved and recommended by the user B. It becomes impossible. In this case, the requesting user A is selected from the recommendation requesting user (user G) remaining on the screen corresponding to the screen G20. The user G requested to be recommended is urged to recommend from the users J and K by the screen corresponding to the screen G21.

  On the system side (reliability calculation device 10 side), the matching unit 14 may store a history of people who use the ride share and use it for calculation of usage charges. In this history, the place of riding (GPS), the place of getting off (GPS), the intermediary friend (if there is an intermediary), and the friend of the friend who got in the car are recorded. A fee may be calculated based on the usage frequency and distance, and a certain percentage of the usage fee may be fed back to the intermediary friend. The usage fee (and feedback fee) is calculated in advance based on the acquisition conditions in the condition acquisition unit 141 and estimated values for ride sharing, and both the requesting side and the licensing side (and brokerage) May be presented for confirmation.

  Hereinafter, supplementary items will be described.

  As a first supplementary matter, in the description of the present invention, “connection” has been described as a friend relationship, but since it is based on the registration relationship of addresses, the “connection” is an acquaintance relationship or family relationship. Others may be used.

  As a second supplementary matter, the matching by the matching unit 14 explained the matching between one user A who requests to ride the vehicle and one user who provides the vehicle, taking the case of vehicle riding as an example. There may be two or more users who request to get on the vehicle.

If there are two or more requesters, the candidate presentation unit 142 confirms the following conditions (1) to (3).
(1) All conditions on the request side are input to the condition acquisition unit 141 so that the conditions (including the date and time of sharing) set for the other party match.
(2) There is a friend who is common to all requesters, and the friend of the friend can satisfy the above condition and provide the vehicle.
(3) Any two parties selected from the provider of the vehicle and all of the requesting parties are not direct friends.

  After confirming all the above conditions (1) to (3), the confirmation unit 143 may mediate the final agreement between the provider of the vehicle and all of the requesting parties.

  Alternatively, one user A who requests to ride a vehicle is a representative of a plurality of requesters including acquaintances of the A, and after setting the condition acquisition unit 141 that the plurality requests the boarding, According to the embodiment of one user A described above, one vehicle provider may be found and a plurality of people may be realized.

  Alternatively, the vehicle is not provided by the user but by a separate trader or the like, and in order to set the riding conditions at that time, matching may be performed among the users.

  As a third supplementary matter, when the user attribute (gender, age, address, presence / absence of driver's license, etc.) is acquired by the condition acquisition unit 141, the user uses a predetermined mathematical formula based on a match or mismatch of a predetermined item in the user attribute. The compatibility between them may be quantified. In the candidate presentation unit 142, a value obtained by further adding the compatibility to the reliability calculated by the reliability calculation unit 13 may be used instead of the reliability described above.

  As a fourth supplementary matter, the reliability calculation device 10 may be one in which the matching unit 14 is omitted from the configuration of FIG. Depending on the reliability calculated by the reliability calculation unit 13, matching may be realized by other methods, or the reliability may be used for purposes other than matching. In a configuration in which the matching unit 14 is not omitted, the reliability calculation device 10 may be referred to as the matching device 10.

  DESCRIPTION OF SYMBOLS 10 ... Reliability calculation apparatus, 11 ... Address acquisition part, 12 ... User map production | generation part, 13 ... Reliability calculation part, 14 ... Matching part, 141 ... Condition acquisition part, 142 ... Candidate presentation part, 143 ... Determination part

Claims (12)

A reliability calculation device that calculates the reliability of other users in each user via each user's terminal,
From each terminal, an address acquisition unit that acquires information on the address of the registered other user's terminal,
Using the acquired address, each user as a node, a user map generation unit that generates a user map as a graph with an edge between users with address registration,
A reliability calculation apparatus comprising: a reliability calculation unit that calculates the reliability of other users in each user based on a path from the user to the other user using the generated user map.   The reliability calculation apparatus according to claim 1, wherein the reliability calculation unit calculates the reliability based on the number of paths from each user to another user and the number of steps of each path.   The reliability calculation unit considers the reliability of other users in each user based on the path from the user to the other user and the update frequency of the address registered in the user's terminal existing on the path. The reliability calculation device according to claim 1, wherein the reliability is calculated.   The said reliability calculation part distinguishes and calculates the reliability of the other user in each user for every shortest step number of the path | pass from the said user to another user, The calculation of any one of Claim 1 thru | or 3 characterized by the above-mentioned. Reliability calculation device.   The reliability calculation unit calculates the reliability of other users in each user in order from the side with the smallest number of shortest steps, and based on the path from the user to the other user, for users existing on the path The reliability calculation apparatus according to claim 4, wherein the reliability is calculated in consideration of the reliability already calculated.   The reliability calculation unit determines the reliability of other users in each user based on the path from the user to the other user, taking into account the communication history between the terminals of the users existing on the path, The reliability calculation apparatus according to claim 1, wherein the reliability calculation apparatus calculates the reliability.   7. The reliability calculation apparatus according to claim 1, wherein the reliability relates to sharing a predetermined event between users.   A matching unit that mediates matching between terminals by presenting matching candidate users to the requesting user based on the calculated reliability to the requesting user's terminal. The reliability calculation apparatus according to claim 1, wherein The matching unit is
A condition acquisition unit that acquires a condition for matching from a predetermined item from each user's terminal; and
A candidate presenting unit that notifies the requesting user's terminal of a predetermined number of users on the side where the acquired condition is met with the requesting user and the calculated reliability is high;
9. The method according to claim 8, further comprising: a user terminal selected from the matching candidates by the requesting user; and a determination unit that performs mediation for determining matching between the requesting user terminal. The reliability calculation device described.   The reliability calculation apparatus according to claim 9, wherein the candidate presentation unit selects a matching candidate from users connected to the requesting user in two or more steps with reference to the user map.   The reliability calculation apparatus according to claim 9, wherein the condition acquisition unit further acquires position information from a terminal of each user as a condition for the matching.   The reliability calculation apparatus according to claim 9, wherein the matching relates to a vehicle ride between users.