JP6653897B1 - Information processing device for matching advisors - Google Patents

Abstract

An information processing apparatus that presents an optimal advisor candidate list to a user is provided. In an information processing apparatus, means for receiving a question input by a user, means for calculating an advisor's score based on at least the question and an attribute of the advisor, and selecting an advisor based on the calculated score Means, means for receiving a response to the question entered by the selected advisor, means for displaying a list of advisor candidates based on the received response, and an identifier of the advisor selected by the user from the displayed candidate list. , A means for realizing a session with the selected advisor, and means for calculating an incentive point based at least on the performance of the session. [Selection diagram] Fig. 1

Description

  The present invention relates to an information processing apparatus for matching an advisor (expert).

  A method of matching an advisor for solving a consultation from a user (client) has been conventionally proposed in various forms.

  Patent Literature 1 presents a list of advisors based on a keyword input by a user (a requester). However, even if an advisor is selected based on a keyword, there is often a mismatch, and an optimal advisor is presented. I can't say that.

  Non-Patent Document 1 is a service for matching an advisor, but performs matching artificially based on the knowledge of an intermediary himself / herself that mediates between the user (client) and the advisor, so that the time and time required for the matching are high. The cost is very high, and the quality of the matching quality that is unique to the intermediary is also caused. Therefore, there is a case where no merit is felt for the cost paid.

JP 2017-73007 A

https: // glg. it / ja /

In addition to the above-described conventional techniques, there are various problems in matching an advisor.
For example, since the obsolescence of information is accelerated due to the expansion of business and technical fields and the acceleration of progress, there is a problem that the degree of difficulty in registering an optimal advisor in its own database continues to increase.
Furthermore, even if there is no specific "question" of the client, recruiting only through the Internet "to gather as many specialists as possible" will result in out-of-focus specialists matching There is also a problem that it is easy to be done.
In the first place, there is a problem that a certain number of specialists do not respond to a registration request from a matching service provider.

  According to one embodiment of the present invention, it is an object of the present invention to provide an information processing device that presents an optimal advisor candidate list to a user.

  According to another aspect of the present invention, it is an object of the present invention to provide an information processing apparatus that presents a more optimal advisor candidate list through introduction from an advisor.

According to one aspect of the invention,
Means for receiving the question entered by the user;
Means for calculating an advisor's score based on at least the question and the advisor's attributes;
Means for selecting an advisor based on the calculated score;
Means for receiving a response to the question entered by the selected advisor,
Means for displaying an advisor candidate list based on the received response;
Means for receiving an identifier of an advisor selected by the user from the displayed candidate list;
Means for realizing a session with the selected advisor,
Means for calculating an incentive point based at least on the performance of the session;
An information processing apparatus characterized by comprising:

  According to one aspect of the present embodiment, it is possible to provide a user with an optimal advisor candidate list. In addition, through referrals and incentives, excellent advisors can be gathered. Further, even if the specialist who does not respond to the registration request from the provider of the matching service can increase the probability of responding if the introduction is from a friend who is a friend, the quality of the service can be improved.

  Other objects, features and advantages of the present invention will become apparent from the following description of embodiments of the present invention with reference to the accompanying drawings.

1 shows a block diagram of an information processing apparatus according to one embodiment of the present invention. 4 shows a flowchart of information processing according to an embodiment of the present invention. 1 conceptually shows an overall view of an information processing system including an information processing apparatus according to an embodiment of the present invention. 4 shows a data structure according to an embodiment of the present invention.

  FIG. 3 conceptually shows an overall view of an information processing system 3000 including an information processing apparatus according to an embodiment of the present invention.

  In the information processing system 3000 of this embodiment, an information processing apparatus 1000, a user terminal 3100, and an advisor terminal 3200 (3200-1, 3200-2,...) Are connected via a network 3300.

  An example of a block diagram of the information processing apparatus 1000 is shown in FIG. Details will be described later.

  The user terminal 3100 is a terminal (for example, a smartphone or a desktop personal computer) used by a user (client). The service provided by the information processing apparatus 1000 can be logged in using an identifier (ID) and a password assigned to each user.

  The advisor terminal 3200 is a terminal used by an advisor (expert). The service provided by the information processing apparatus 1000 can be logged in using an identifier (ID) and a password assigned to each advisor.

  Network 3300 may be a wireless and / or wired leased line.

  FIG. 1 shows a block diagram of an information processing apparatus 1000 according to one embodiment of the present invention. The information processing apparatus 1000 of this embodiment includes an output (screen display) unit 1110, an input unit 1120, an input analysis unit 1130, a score calculation unit 1140, an advisor selection unit 1150, a user management unit 1160, an advisor management unit 1170, and an advisor management DB 1210. , An advisor attribute DB 1220, an advisor score DB 1230, a user management DB 1240, an advisor evaluation DB 1250, an introduction history DB 1260, a control unit (processor) 1310, and an interface unit 1320.

  An output (screen display) unit 1110 performs processing related to output. The output processing includes displaying an output result on a screen.

  The input unit 1120 performs processing related to input.

  The input analysis unit 1130 analyzes data processed by the input unit.

  The score calculation unit 1140 calculates the score of the advisor based on the question from the user (client) and the attribute of the advisor.

  The advisor selecting unit 1150 selects an advisor.

  The user management unit 1160 manages user (client) IDs and inquiry histories, and performs user ID collation and processing of inquiry data.

  The advisor management unit 1170 manages the ID of the advisor (expert) and the history of inquiry responses, and performs collation of user IDs and processing of inquiry data.

  The advisor management DB 1210 stores an advisor's ID and a history of inquiry responses. FIG. 4A shows an example of the data structure.

  The advisor attribute DB 1220 stores so-called profiles of the attributes of the advisor, such as, for example, career, educational background, and a field in which advice can be given (area with knowledge). FIG. 4B shows an example of the data structure.

  The advisor score DB 1230 stores a score (evaluation value) calculated based on the attribute of the advisor. Further, when the score is corrected after the fact by the evaluation from the user and the learning of the model, the corrected score is also stored. Further, a score correction history (a time at which the score was corrected or a corrected score amount) may be stored. FIG. 4C shows an example of the data structure.

  The user management DB 1240 stores a user ID and an inquiry history. FIG. 4D shows an example of the data structure.

  The advisor evaluation DB 1250 stores the evaluation from the user to the advisor after the session between the user and the advisor has ended. FIG. 4E shows an example of the data structure.

  The introduction history DB 1260 stores an identifier of an advisor in association with an identifier of another advisor introduced by the advisor and a consultation result of the advisor introduced. Further, an introduction date may be stored. An example of the data structure is shown in FIG.

  The incentive point DB 1270 stores points based on the results of the advisor. FIG. 4G shows an example of the data structure. Further, the reason for giving points may be stored. Further, a point occurrence date as shown in FIG. 4 (g) may be stored in association with the consultation date of FIG. 4 (d) or the introduction date of FIG. 4 (f).

  The control unit (processor) 1310 controls each unit and each DB inside the information processing apparatus 1000.

  The interface unit 1320 transmits and receives data to and from a device external to the information processing device 1000 via the network in FIG.

  FIG. 2 shows a flowchart of information processing according to one embodiment of the present invention.

  In S2010, the user inputs a matter to be consulted with the advisor (hereinafter, “question”). A question point of the user is input into a predetermined input item. Here, the input item may be a single text box, but a plurality of text boxes may be provided so that the "question" can be easily analyzed.

  In step S2020, the server (for example, the information processing apparatus 1000) analyzes the input items in response to the input “question”, and obtains the attribute information and the past information of each advisor registered in the advisor database in advance from the trained model. The score is calculated based on the information such as the corrected score and the activity history. In generating a score, an AI / statistical machine learning approach centering on natural language processing is used. As another example, a traditional statistical method may be used, and a score may be generated using a determination table. There are a plurality of data items to be input by the user, for example, an item for inputting a field to be inquired, an item for actually asking a question, and the like. The server, for example, extracts a keyword for each input item and weights the keyword according to the input item (for example, a keyword input to a field item has a higher importance level). And keywords having more specific expertise may be judged to have higher importance), so that an optimal advisor can be searched.

  In S2030, the server generates an advisor candidate list (s) based on the calculated scores. For example, a candidate list of the top five advisors having a high calculated score is generated. Further, based on the attribute information of the advisor, if there is a possibility that a conflict of interest may occur due to the answer to the “question”, the server may exclude the conflict from the advisor candidate list. The server transmits the input “question” to each of the advisors on the candidate list to the advisor terminal.

  As another example of S2020 and S2030, an advisor may be searched under a plurality of conditions in order to avoid filling the candidate list with people of similar carriers. For example, in the search for an advisor in S2020, a search is performed under a certain search condition to select the top one advisor, and then the search condition is changed by changing the weighting of the keyword or the like, and another search is performed. A mechanism for selecting one advisor may be adopted. Then, in S2030, a candidate list combining search results under a plurality of conditions in S2020 may be created. As a result, a candidate list that takes into account the diversity of the advisors can be generated, so that the possibility of being introduced through more various routes is increased.

  As yet another example, the search condition of S2020 may be configured to search not only the ability to answer “question” but also an advisor that can introduce another advisor who may have such ability. . For example, the introduction results in the introduction history DB of FIG. 4F may be quantified so that a search can be performed based on the results of the introduced advisors.

  In S2040, the selected advisor returns a response to the “question” from the advisor terminal to the server. If the answer is possible, the selected advisor also answers the answer (input in a predetermined item). If the server receives a reply indicating that the answer is not available, the corresponding advisor may be deleted from the candidate list, and the score of the deleted advisor may be corrected (for example, corrected downward).

  In S2050, at the same time, the server asks the selected advisor whether or not another advisor can be introduced. Here, another advisor means an advisor who seems to be more appropriate (has detailed knowledge) than himself / herself for "question". The selected advisor inputs an identifier (e.g., user ID, mail address) of another advisor that can be introduced from the terminal of the selected advisor (e.g., the advisor terminal 3200-1 in FIG. 3). It may be determined that the referral procedure has been performed. The server transmits an “inquiry” or the like to another advisor based on the input identifier of another advisor (for example, to the advisor terminal 3200-2 in FIG. 3).

  In S2060, another advisor returns whether or not the answer to the “question” is possible. If another advisor can introduce another advisor, the same information processing as in S2050 may be performed. However, the introduced advisor may be configured so that it cannot introduce further advisors, and the advisor is configured so as to limit the number of referral chains (such as an advisor introducing another advisor). May be.

  In S2070, it may be executed simultaneously with S2060. If another introduced advisor is not registered in the database of the server, the other advisor is prompted to input attribute information (profile) and the like, and a new ID and password are input. Issue. Further, a score is calculated based on the input attribute information. Furthermore, the score may be revised upward in consideration of the point at which the introduction was made. Then, when there is another introduced adviser, the same processing as that of S2050 is performed on the adviser.

In S2080, the server makes a list of candidates for the advisor, including a reply from the advisor (a response to the “question” and a response), and transmits the list to the user terminal. The user terminal displays the transmitted candidate list on the screen. User refers to the candidate list displayed, selects an ad bar Isa one person or plural persons.

  In S2090, the user has a session (consultation) with the selected advisor.

In S2100, after the session ends, the user evaluates the advisor. The evaluation method may be arbitrary. For example, a predetermined item may be answered with YES / NO or a score may be given. Then, the server stores the evaluation to the ad server organizer to Advisor evaluation DB. Furthermore, by learning the sessions that were actually matched and their evaluations, the logic itself was updated to more accurate advisors that could be introduced and scoring that gave higher scores to the advisors that could be answered. You may.

In addition, the current session, if based on the advisors that was introduced in the introducer is to introduce the history DB, and the identifier of the advisor that was introduced, the string with stores and Advisor of the identifier that has been introduced. As a result, the number of advisers introduced by the advisers can be calculated. Further, the introduction result status is also stored in association with the status. Here, the referral result status indicates that "consultation has been conducted" indicating that the session (consultation) was actually conducted with the introduced advisor, or that the user was actually introduced but was not selected by the user. There is "consultation failure". Thus, the number of consultations with the actually introduced advisor can be calculated. By adopting such a status, it is possible to evaluate the performance of introducing an advisor.

  Note that the advisor can also evaluate the user.

  In S2110, an incentive point is calculated based on the implementation of the session and the evaluation. For example, if the (first selected) advisor introduces another advisor and the introduced advisor conducts a session (consultation), +250 points are added to the advisor who introduced the advisor. At the same time, +400 points are added to the introduced advisor. Further, the incentive points may be added or subtracted based on the evaluation from the user in S2100. Further, a reward amount to the advisor may be calculated based on the incentive points. Further, the point giving date may be stored.

  Although the present embodiment has been described with reference to functional blocks and flowcharts, another embodiment is directed to hardware resources of so-called five major computers (control device, arithmetic device, storage device, input device, output device). Alternatively, it can be realized by cooperating software.

  Although the embodiments of the present invention have been described above, various alternatives, modifications, or variations are possible for those skilled in the art based on the above description, and the present invention is not limited to the above various embodiments without departing from the spirit thereof. It is intended to cover alternatives, modifications or variations.

Claims (3)

Means for receiving the question entered by the user;
Means for calculating an advisor's score based on at least the question and the advisor's attributes;
Means for selecting an advisor based on the calculated score;
Means for receiving a reply to the questions which the selected the advisor is input,
A means for displaying a candidate list from which an advisor who has rejected the reply is deleted based on the received reply,
Means for receiving an identifier of an advisor selected by the user from the displayed candidate list;
Means for calculating an incentive point based at least on conducting a session with the selected advisor ;
Equipped with a,
The means for selecting the advisor, when the selected advisor inputs an identifier of another advisor, further selects the other advisor ,
It said means for calculating the incentive points, further characterized by computing the incentive points to the selected the advisors and the another advisor, the information processing apparatus. An identifier that was introduced advisor, further comprising the means for storing in association with the identifier and Introduction actual status of introduction has been advisor, the information processing apparatus according to claim 1. Further comprising means for storing an evaluation to an advisor in the session;
The information processing apparatus according to claim 1 , wherein the means for calculating the score of the advisor calculates the score based on the evaluation.