Trust relationship estimation device, trust relationship estimation method and computer program

Abstract

To provide a trust relationship estimation device for estimating strength of a trust relationship built between lecturers and students.SOLUTION: A trust relationship estimation device 1 for estimating a trust relationship built between lecturers and students, includes an information acquisition unit 11 for acquiring transmissions by a lecturer and a student, and a trust relationship analysis unit 12 for, referring to the transmissions, when detecting that the lecturer and the student have alternately transmitted information of the same topic at least three times or more, estimating that a trust relationship is built between the lecturer and the student through the transmissions, and storing it in a storage unit 50 as information relating to the trust relationship.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a trust relationship estimation device, a trust relationship estimation method, and a computer program. [Background technology]

[0002] It is common for instructors to provide training to students. Instructors and students build a relationship of trust through communication. The stronger this relationship of trust is, the more students understand the instructor's personality and teaching methods, and the more they can fully absorb the instructor's instruction. However, students who are closed off will not be able to receive even the best instruction. In addition, the stronger the relationship of trust is with students, the more effective the instructor can provide instruction that is tailored to the students, and the greater the training effect on students. In this way, even if the training content, instructor, and students are the same, the training effect will vary greatly depending on the state of the relationship of trust between the instructor and the students. In other words, in order to measure and estimate the essential training effect, it is essential to consider the relationship of trust, and a technology is needed to estimate the strength of the relationship of trust built between the instructor and the students.

[0003] Patent Document 1 describes an information processing device that analyzes the tendency of mutual comments between multiple students and presents a comment map showing the student community. Patent Document 2 discloses a user relationship calculation device that calculates and outputs the relationship between users based on the number and size of sent e-mails received from a mail analysis unit 110. Patent Document 3 describes a personal relationship information display device that weights the strength of relationships between people by using metadata such as schedules. Patent Document 4 describes a personal relationship estimation device that estimates the personal relationships between users based on how frequently each user communicates and in which areas they were present at the same time. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Patent Publication No. 2022-116607 [Patent Document 2] JP 2006-260099 A [Patent Document 3] Patent Publication No. 2007-193685 [Patent Document 4] Patent Publication No. 2010-165097

[0005] However, none of the studies were able to adequately estimate the level of trust between the instructor and the trainee that would be necessary for proper guidance to be provided. Summary of the Invention [Problem to be solved by the invention]

[0006] Therefore, an object of the present invention is to provide a trust relationship estimation device that can estimate the strength of the trust relationship built between a lecturer and a student. [Means for solving the problem]

[0007] In order to achieve the above-mentioned object, a trust estimation device according to one aspect of the present invention is a trust estimation device that estimates a trust relationship established between a lecturer and a student, and includes an information acquisition unit that acquires communications by the lecturer and the student, and a trust relationship analysis unit that, by referring to the communications, detects an event in which the lecturer and the student have made communications alternately at least three times on the same topic, estimates that the trust relationship has been established between the lecturer and the student through the communications, and stores this information regarding the trust relationship in a memory unit.

[0008] The trust relationship analysis unit may acquire the communications by the instructor and the student as audio data or text data, and analyze the content of the communications to infer that the communications by the instructor and the student are related to the same topic.

[0009] The information acquisition unit may acquire messages from the instructor and the students as text data via a terminal connected to the trust relationship estimation device, and the trust relationship analysis unit may infer that the messages, which have been linked and input in advance, are related to the same topic.

[0010] The memory unit may store the technical level of the instructor and the technical level of the student in a specified technical field, and the trust relationship analysis unit may vary the degree of trust relationship that is estimated to have increased between the instructor and the student in the event depending on the technical level of the student in which the event occurred, the technical level of the instructor, or the difference in the technical levels of the student and the instructor.

[0011] The memory unit stores information about a training session in which the instructor and the students participate, and the trust relationship analysis unit may refer to the memory unit to identify identification information of the training session in which the event occurred, and vary the degree of trust relationship that is estimated to have increased between the instructor and the students during the event depending on the number of people simultaneously participating in the training session.

[0012] The trust relationship analysis unit may be configured to decrease the degree of the trust relationship between the instructor and the student as time passes since the event that established the trust relationship between the instructor and the student.

[0013] The information on the trust relationship stored in the storage unit may be issued for each of the students.

[0014] In order to achieve the above-mentioned object, a trust relationship estimation method according to another aspect of the present invention is a trust relationship estimation method for estimating a trust relationship established between a lecturer and a student, which includes an information acquisition step of acquiring communications from the lecturer and the student, and a trust relationship analysis step of referring to the communications and, upon detecting an event in which the lecturer and the student have alternately transmitted messages on the same topic at least three times, inferring that the trust relationship has been established between the lecturer and the student through the communications, and storing this information regarding the trust relationship in a memory unit, which is executed by a computer.

[0015] In order to achieve the above-mentioned object, a computer program according to yet another aspect of the present invention is a computer program for estimating a relationship of trust established between a lecturer and a student, which causes a computer to execute an information acquisition step of acquiring communications from the lecturer and the student, and a trust relationship analysis step of, when the communications are referenced and an event is detected in which the lecturer and the student have alternately made communications on the same topic at least three times, estimating that a relationship of trust has been established between the lecturer and the student through the communications, and storing this information about the trust relationship in a memory unit.

[0016] The computer program can be provided by downloading via a network such as the Internet, or can be provided by recording it on various computer-readable recording media. Effect of the Invention

[0017] According to the present invention, it is possible to estimate the strength of the trust relationship built between the instructor and the students. [Brief description of the drawings]

[0018] [Figure 1] 1 is a functional block diagram showing functions of a trust relationship estimation device according to a first embodiment of the present invention, and each component connected via the trust relationship estimation device and a network. FIG. [Diagram 2]FIG. 11 shows examples of data tables stored in the trust relationship estimation device, including (a) a student table storing information about students, (b) a student level table storing information about the skill level of students, (c) a lecturer table storing information about lecturers, and (d) a lecturer level table storing information about the skill level of lecturers. [Diagram 3] 5A and 5B are diagrams showing examples of data tables stored in the trust relationship estimation device, including (a) a training menu table storing information about a training menu, and (b) a training table storing information about training to be implemented. [Figure 4] FIG. 11 shows examples of data tables stored in the trust relationship estimation device, including (a) a trust relationship event table that stores information on events in which trust relationships are built, and (b) a trust relationship index table that stores a trust relationship index indicating the trust relationship built between a student and an instructor. [Diagram 5] 13 is an example of a screen displayed on a terminal connected to the trust relationship estimation device. [Figure 6] 5A and 5B are sequence diagrams showing a flow of processing by the trust relationship estimation device, in which (a) is a first example and (b) is a second example. FIG. [Figure 7] 5A and 5B are sequence diagrams showing a flow of processing by the trust relationship estimation device, in which (a) is a third example and (b) is a fourth example. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] Hereinafter, an embodiment of a trust relationship estimation device according to the present invention will be described with reference to the drawings. The trust relationship estimation device is a device that estimates a trust relationship built between a lecturer and a student by referring to information exchanged between the lecturer and the student. The trust relationship estimation device displays information related to the trust relationship on an appropriate terminal connected via a network or the like.

[0020] As shown in Fig. 1, the trust estimation device 1 is configured to be able to communicate with a plurality of terminals 2 via a network NW. The trust estimation device 1 is, for example, a server. In this embodiment, the mutual communication between the trust estimation device 1 and each terminal 2 is wireless, but some or all of the connections may be wired. Also, while there are two terminals 2 in Fig. 1, there may be any number of terminals. A trust estimation system 100 may be configured by the trust estimation device 1 and one or more terminals 2.

[0021] Terminal 2 Terminal 2 is a terminal that accepts input of information about the instructor who conducts the training and the students who receive the training. Instructor terminal 2a and student terminal 2b are examples of terminal 2, and when the instructor terminal 2a and the student terminal 2b are described without distinguishing between them, they may be collectively referred to as terminal 2. In the following description, as an example, instructor terminal 2a is a terminal operated by an instructor who holds the training. Student terminal 2b is described as a terminal operated by a student of the training. The instructor terminal 2a and the student terminal 2b are described by separating their functions for convenience only, and training can be received from instructor terminal 2a, and training and instruction can be provided from student terminal 2b.

[0022] The terminal 2 is, for example, a smartphone, a tablet terminal, or a computer. The terminal 2 may be a device worn by a wearer, such as so-called smart glasses, a headset, or a smart watch. The terminal 2 is configured as a functional block consisting of an input unit 21, an output unit 22, and a communication processing unit 23, using a CPU (Central Processing Unit), a computer program executed by the CPU, a RAM (Random Access Memory) or a ROM (Read Only Memory) for storing the computer program and predetermined data, and the like. The input units 21a and 21b, the output units 22a and 22b, and the communication processing units 23a and 23b are examples of the input unit 21, the output unit 22, and the communication processing unit 23, respectively, and when describing without distinguishing between them, these may be collectively referred to simply as the input unit 21, the output unit 22, and the communication processing unit 23, respectively.

[0023] The input unit 21 is a component for inputting data, and may be realized by a touch panel or the like, or may have physical keys such as a keyboard. The input unit 21 also has a microphone for acquiring audio, and a camera for acquiring still or moving images. The instructor terminal 2a accepts inputs, for example, from the instructor regarding instruction to the students in the form of text data, audio, video, etc. The student terminal 2b accepts messages to the instructor and responses to the instructor's instruction in the form of text data, audio, video, etc.

[0024] Furthermore, the input unit 21 may receive an input of more direct information regarding the trust relationship. For example, the input unit 21a may receive a record of an event in which a trust relationship with a student was built by the instructor. The input unit 21a may also receive an input from the instructor to the effect that a trust relationship with a student has been built, or evaluation information on the student's response, etc. The input unit 21b of the student terminal 2b may receive evaluation information on the contents of the instructor's response, etc., from the student. The evaluation information may be input by displaying options on the screen of the terminal 2 and selectively input by the student or the instructor. For example, this information is not transmitted to other terminals 2 and is referred to in the estimation of the trust relationship by the trust relationship estimation device 1.

[0025] The input unit 21 may include an appropriate sensor for detecting information about a user of the terminal. The input unit 21 may also acquire location information. The input unit 21 may have a beacon function for acquiring information about other nearby terminals.

[0026] The output unit 22 is configured to output information related to the training, and is realized, for example, by a display that outputs video and a speaker that outputs audio. The display may form a touch panel display together with the touch panel of the input unit 21. The output unit 22, for example, displays video of the instructor or the students and plays back the voices of the instructor or the students.

[0027] The instructor and the students can remotely converse with each other and exchange text data using the input units 21a, 21b and output units 22a, 22b of the terminals 2a, 2b.

[0028] Instead of each person using a different terminal 2a, 2b, a single terminal 2 may be used to record or collect audio of both the instructor and the students. In this case, data acquired by a single input unit may be processed by the trust relationship analysis unit 12 to separate messages from the instructor and messages from the students. A single terminal 2 may be provided with multiple input units that acquire messages from the instructor and the students, respectively.

[0029] The communication processing unit 23 is a processing unit that can execute data transmission and reception processing with the trust relationship estimation device 1 according to a predetermined protocol via a network NW such as the Internet, and is realized by an application, a Web browser, or the like. The communication processing unit 23a of the instructor terminal 2a transmits the instructor's voice and video input to the input unit 21a of the instructor terminal 2a to the student terminal 2b, and receives the students' voice and video input to the student terminal 2b. The communication processing unit 23b of the student terminal 2b transmits information input to the input unit 21b of the student terminal 2b to the instructor terminal 2a, and receives the students' voice and video input to the instructor terminal 2a.

[0030] The communication processing unit 23 may also receive an input form or the like that allows input of information as a transmission from the trust relationship estimation device 1 or an appropriate external device. The communication processing unit 23 may also transmit information input into the input form to the trust relationship estimation device 1 or another terminal 2.

[0031] The communication processing unit 23 also transmits the messages from the instructor and the students acquired by the input unit 21 to the trust relationship estimation device 1. That is, for example, the communication processing unit 23a transmits the message from the instructor acquired by the input unit 21a of the instructor terminal 2a together with the identification information of the instructor terminal 2a to the trust relationship estimation device 1. The communication processing unit 23a may transmit the identification information of the student terminal 2b that is permitted to receive the message from the instructor together with the message from the instructor to the trust relationship estimation device 1. The communication processing unit 23b transmits the message from the instructor acquired by the input unit 21b of the student terminal 2b together with the identification information of the instructor terminal 2a to the trust relationship estimation device 1. The communication processing unit 23b may transmit the identification information of the instructor terminal 2a that received the message from the student together with the message from the student to the trust relationship estimation device 1. As a result, the trust relationship estimation device 1 acquires the identification information and message content of the instructor and student who have made messages to each other.

[0032] ● Trust relationship estimation device 1 The trust relationship estimation device 1 is configured with a CPU (Central Processing Unit), a computer program executed by the CPU, a RAM (Random Access Memory) and a ROM (Read Only Memory) for storing the computer program and predetermined data, and is configured with functional blocks mainly consisting of a memory unit 50, an information acquisition unit 11, a trust relationship analysis unit 12, and a communication processing unit 13, as shown in FIG. 1.

[0033] The functional units of the trust relationship estimation device 1 may be realized by a plurality of hardware configurations, or may be partially or entirely configured by cloud computing. Also, some of the functional units of the trust relationship estimation device 1 may be provided in the instructor terminal 2a or the student terminal 2b.

[0034] The storage unit 50 is a functional unit that stores information about the instructors and students involved in the training. The storage unit 50 stores, for example, an instructor information database (in the following description, "database" will be referred to as "DB") 51 that stores information about the students, an instructor information DB 52 that stores information about the instructors, a training information DB 53 that stores information about the training, and a trust relationship information DB 54 that stores information about the trust relationship built between the instructor and the students.

[0035] An example of a table included in each DB of the storage unit 50 will be described with reference to Figs. 2 to 4. The table configuration is an example, and some of the tables described below may be integrated into one table, or one table may be composed of multiple tables. In addition, although a so-called relational database in a tabular format is described as an example in this specification, the technical scope of the present invention is not limited to this, and other formats such as a graph database may also be used. According to a configuration in which information is stored in a graph database, information can be stored accurately and concisely even in a case where the same person is a participant in one training course and a lecturer in another training course, and a network-like relationship structure such as a social network service (SNS) can be easily managed.

[0036] Fig. 2(a) is an example of a student table T1 included in the student information DB 51. In the student table T1, for example, the name, the company, and contact information are stored in association with the identification information of the student. Also, Fig. 2(b) is an example of a student level table T2 included in the student information DB 51, which stores the technical level of the student. In the student level table T2, for example, the technical level of the student is stored for each technical field in association with the identification information of the student.

[0037] Fig. 2(c) is an example of a lecturer table T3 included in the lecturer information DB 52. The lecturer table T3 stores, for example, the name and contact information of the lecturer, the training menu that the lecturer can teach, etc., in association with the lecturer's identification information. Fig. 2(d) is an example of a lecturer level table T4 included in the lecturer information DB 52 and storing the lecturer's technical level. The lecturer level table T4 stores, for example, the lecturer's technical level for each technical field in association with the lecturer's identification information.

[0038] 3(a) is an example of a training menu information table T5 included in the training information DB 53. The training menu is information indicating the nature of the training itself, such as the technical field of the training and the required number of days. For example, the training menu information table T5 stores the identification information of the training menu, the technical field, the technical level, the number of training days, etc. in association with each other.

[0039] FIG. 3(b) is an example of a training information table T6 that stores information about each training session implemented according to a certain training menu. In the training information table T6, identification information of the training menu, identification information of the instructor in charge, implementation schedule, and the number of participating students are stored in association with each other. The number of participating students may be the planned number of students, or the actual number may be stored after the training session. In addition, the number of participating students may be corrected to the actual number after the training session is held after the planned number of students is stored. If the training session is held over multiple days or multiple hours, the actual number of participating students for each day or hour may be stored. In addition, the training information table T6 stores identification information of the participating students in association with identification information of the training session.

[0040] 4(a) is an example of a table stored in the trust relationship information DB 54, which is an example of a trust relationship event table T7 that accumulates and stores events related to the building of a trust relationship that occurred between a student and a lecturer. The trust relationship event table T7 stores, for example, identification information of the student and lecturer who built a trust relationship, the time when the event occurred, identification information of the training held in which the event occurred, and the mode of the transmission route from the student to the lecturer or from the lecturer to the student. The mode of the transmission route includes, for example, information such as an online lecture, daily report, email, chat, and face-to-face lecture. The mode of the transmission route also includes information on whether the transmission originated from the student or the lecturer.

[0041] The trust relationship event table T7 stores multiple messages about the same topic as one event. In the following explanation, the sender who first sends a message about a topic is called the first sender, and the sender who replies to the first sender about the topic is called the second sender. Both the first sender and the second sender can be either a lecturer or a student.

[0042] For example, in event No. 001, a student S01 asks a question as a first sender in an online lecture (first route), a lecturer T01 answers the question as a second sender in the same lecture (second route), and immediately afterwards, the student S01, the first sender, responds (third route). Event No. 002 is an example in which, in an operation of inputting a daily report on a day of training, a message by text data entered in the daily report is integrated and stored as one event. Event No. 003 is an example in which a lecturer answers a question from a student in an online lecture by email after the lecture, and the student sends a response by chat on another day. In this way, in the present invention, such exchanges on the same topic are estimated and stored as one event in which a trust relationship is built. The process of estimating that the topics of messages are the same will be described later. In addition, although the figure shows up to the third route, four or more detected routes may be stored. The reason why the present invention detects the establishment of a trust relationship by three or more routes will be described later.

[0043] 4(b) is an example of a table stored in the trust relationship information DB 54, and is an example of a trust relationship index table T8 in which a trust relationship index indicating the trust relationship established between the students and the instructor is stored. The trust relationship index table T8 stores identification information of the students and the instructor in association with the trust relationship index between the students and the instructor. The trust relationship index table T8 may be linked to an event number in the trust relationship event table T7 that contributes to the trust relationship between the students and the instructor.

[0044] Return to FIG. 1. The information acquisition unit 11 is a functional unit that acquires transmissions by instructors and students. The information acquisition unit 11 acquires input information from the input unit 21 of the terminal 2. The input information is, for example, information to be transmitted to another terminal 2, and may be voice, video, or text data. The text data may be data written in an input form such as an email, chat, or a web-based questionnaire or daily report, as well as a task shared in a task management tool such as JIRA (registered trademark), or a comment on the task, a program code shared by a software development platform such as GitHub (registered trademark), or a pull request or comment on the program code. The video includes, for example, the reaction of the sender, such as a gesture of nodding or tilting the head, or a facial expression. The input information may also be information selectively input from the terminal 2. The selectively input information may be, for example, an emoticon, a pre-registered illustration, or a so-called stamp.

[0045] The input information may include information that is acquired by the trust relationship estimation device 1 and is not transmitted to the other terminal 2, in addition to the information transmitted to the other terminal 2 as described above. This information is, for example, evaluation information on the students input from the instructor terminal 2a, or evaluation information on the instructor input from the student terminal 2b. The evaluation information is, for example, information that the trust relationship has improved or deteriorated, and the degree of the improvement or deterioration, input by the instructor or student. The evaluation information may be input by selecting an icon displayed on the screen of the terminal 2 or by inputting a numerical value. With this configuration, the actual impression of the instructor or student can be more accurately reflected in the estimation result of the trust relationship.

[0046] The trust relationship analysis unit 12 is a functional unit that estimates the trust relationship built between the instructor and the students.

[0047] The trust relationship analysis unit 12 analyzes the information acquired by the information acquisition unit 11, and identifies the identification information of the lecturer or student who transmitted the information. For example, the trust relationship analysis unit 12 may extract the identification information of the lecturer or student who made the transmission by referring to the identification information of the terminal 2 that accepted the input. For example, in an aspect such as an online lecture where each person transmits from their own terminal 2, the identification information of the lecturer or student can be identified by this method. Also, in an aspect where text data is transmitted and received, such as chat, e-mail, or a web-based input form for daily reports, each person transmits from a terminal 2 linked to their own identification information, so the identification information of the lecturer or student can be identified.

[0048] In the case where the voices of multiple people are collected by one terminal 2, the trust relationship analysis unit 12 may identify the lecturer or the students by analyzing the voices contained in the voice data. In this case, for example, voiceprint data of the lecturer and the students may be stored in advance in the storage unit 50 or an external device, and the lecturer or the students may be identified by comparing the voiceprint data.

[0049] Furthermore, the trust relationship analysis unit 12 refers to the information acquired by the information acquisition unit 11 and identifies the terminal 2 that received each transmission. Through this process, the trust relationship analysis unit 12 can estimate that the instructor and the student are transmitting messages to each other.

[0050] Furthermore, the trust relationship analysis unit 12 detects that a lecturer and a student who are communicating with each other are alternating between communicating with each other about the same topic. The trust relationship analysis unit 12 may acquire messages from the instructor and the students as voice data or text data, and analyze the contents of the messages to estimate that the messages from the instructor and the students are related to the same topic.

[0051] For example, as one mode of analysis, the trust relationship analysis unit 12 may refer to the information acquired by the information acquisition unit 11 and determine whether the messages are about the same topic based on the information included in the message content. For example, in an online lecture, the trust relationship analysis unit 12 may refer to the time information of the message and may presume that the messages are about the same topic if the messages are sent within a predetermined time. In addition, in a daily report input form, messages that are linked to the same day may be presumed to be about the same topic. In addition, messages that are linked and input in advance, not limited to dates, for example, messages posted by selecting the "Reply" button for a specific post on a chat screen, or messages posted in a thread format for a post, may be presumed to be about the same topic. In addition, in the case of email, a message sent by a "Reply" operation in email software may be presumed to be about the same topic as the email that is the reply. For example, the trust relationship analysis unit 12 may use a different method for determining whether the messages are about the same topic depending on the mode of the message.

[0052] The trust relationship analysis unit 12 may use an appropriate machine learning technique to analyze whether the topics are the same. The trust relationship analysis unit 12 may, for example, perform a voice analysis of the voice data to identify the question in the speech section in which the student speaks, and identify the answer in the speech section in which the lecturer speaks following the speech section. The trust relationship analysis unit 12 may also realize a neural network including the contents of the video data, voice data, or text data, the transmission mode, and the transmission time of the data as input data of the input layer. With this configuration, even if the messages are transmitted in different modes, it can be estimated that they are communicating about the same topic, and even in a case such as event No. 003 in the trust relationship event table T7 in FIG. 4(a), it can be detected as an event in which a trust relationship has been established.

[0053] With this configuration, the trust relationship analysis unit 12 can calculate the number of times that the lecturer and the student have exchanged messages on the same topic. When the trust relationship analysis unit 12 detects that messages on the same topic have been exchanged at least three times, it estimates that a trust relationship has been established between the lecturer and the student through the messages.

[0054] The relationship of trust between the instructor and the student is built through communication between the instructor and the student. Since it is communication, the relationship of trust is not built only by the first sender talking unilaterally (the "first route" in the trust event table T7 in Fig. 4(a)), but at least a reaction from the second sender (corresponding to the "second route" in the trust event table T7 in Fig. 4(a). Note that the "simple fact of acceptance or viewing" is also included in the reaction) is a necessary condition. However, from the viewpoint of instruction and the generation of the trust that accompanies it, rather than mere friendship, two mutual transmissions are not a sufficient condition for the establishment. For example, after the instructor instructs the student as the first sender and the student responds as the second sender, the instructor further checks with the student whether "the instruction I gave was understood or received as expected" (the "third route" in the trust event table T7 in Fig. 4(a)).

[0055] In other words, communication in instructional activities is a kind of "feedback." In control engineering, feedback is known to be "the application of an output obtained based on a certain input to the next input." In other words, if the definition of feedback in control engineering is applied to instructional activities, a feedback loop can be said to be established when a first sender makes a further transmission after a second sender transmits a transmission, and this transmission is input to the second sender. According to the trust relationship analysis unit 12, it is estimated that a trust relationship between the instructor and the students is established by the establishment of this feedback loop, and by detecting that alternating transmissions have been made three or more times, it is possible to appropriately estimate whether a trust relationship has been established.

[0056] Furthermore, when a student acts as the first sender of a question to the instructor, they do not simply receive a response from the instructor as the second sender. Rather, they receive the response and express their own understanding to the instructor, for example by restating their understanding in their own words and conveying it to the instructor. Through this, both the instructor and the student realize that mutual instruction has been achieved and a relationship of trust is formed.

[0057] The trust relationship analysis unit 12 may estimate the degree of the established trust relationship by referring to the information acquired by the information acquisition unit 11. The degree of the trust relationship is represented by, for example, a trust relationship index, and the calculated trust relationship index is added to the previous trust relationship index and accumulated.

[0058] The degree of the trust relationship is estimated, for example, by the reply of the second sender on the second path or the reaction of the first sender on the third path. For example, when the first sender asks a question on the first path and the second sender answers on the second path, if the first sender shows a positive reaction, i.e., a reaction that seems convinced, on the third path, it may be estimated that a greater trust relationship has been built than when the first sender shows a negative reaction, i.e., a reaction that seems not convinced. Also, even if only information indicating that the message has been read is transmitted on the third path, rather than a specific message, it may be determined that a certain degree of trust relationship has been built. In this case, the trust relationship index to be increased may be smaller than when a specific message is transmitted.

[0059] The trust relationship analysis unit 12 may analyze and estimate the strength of the logical connection between messages that are estimated to have the same topic using an appropriate artificial intelligence technology, etc., and may increase the trust relationship index less if the logical connection is weak. This is because it is difficult to build a trust relationship in a conversation with a weak logical connection, that is, a conversation where the meanings are not consistent.

[0060] The trust relationship analysis unit 12 may vary the degree of the trust relationship to be built depending on the number of times that messages on the same topic are sent alternately. For example, the trust relationship analysis unit 12 may increase the degree of the trust relationship as the number of messages sent increases.

[0061] The degree of trust that is estimated to have been established due to a certain event may vary depending on the technical level of the trainee, the technical level of the instructor, or the difference between the technical levels of the trainee and the instructor. For example, in a situation where a trainee has a negative reaction, if the trainee's technical level is sufficiently low, there is a high possibility that a trust relationship has not been established. On the other hand, if the trainee's technical level is above a certain level, or if the difference between the trainee's and the instructor's technical level is within a certain level, even if the reaction appears negative on the surface, it may become a constructive dialogue and the trust relationship may increase. Therefore, when a trainee has a negative reaction, the trust relationship analysis unit 12 may estimate that the higher the trainee's technical level is, or the smaller the difference in the technical levels of the trainee and the instructor is, the greater the trust relationship has increased. Since the technical levels of the trainee and the instructor differ depending on the technical field, the trust relationship analysis unit 12 may refer to the technical level in the technical field of the training in which the event in which the alternating transmissions occurred. The trust relationship analysis unit 12 may also estimate the technical field by analyzing the content of the transmission.

[0062] In addition, the degree of trust that is estimated to have been built due to a certain event may differ depending on the trust that was built before the event. If a high level of trust was built before a certain event, there is a high probability that a different reaction will be expressed than when there is relatively little trust. For example, it is considered that a person will react more frankly to a person with whom a high level of trust is built than to a person with whom there is no trust. In other words, for example, even if there is a negative reaction in a relationship with a high level of trust, it may be estimated that a high level of trust has been built compared to a relationship with a relatively low level of trust.

[0063] Furthermore, the trust relationship analysis unit 12 may refer to the training information DB 53, and when an event occurs in which a trust relationship is built between a specific instructor and a trainee, extract information on another trainee who participated in the training in which the event occurred. In this case, the trust relationship analysis unit 12 may also increase the trust relationship between the other trainee and the instructor. This is because the other trainee may strengthen his or her trust in the instructor by indirectly listening to the instructor's instruction. The trust relationship that increases in the other trainee may be smaller than the trust relationship that increases in the trainee who receives direct instruction from the instructor. Also, the trust relationship index that increases in the other trainee may be different depending on the number of people participating in the training at the same time. That is, the smaller the number of people participating at the same time, the larger the trust relationship index that increases. This is because the smaller the number of people in the environment, the easier it is for the instruction given to the other trainee to be heard, which is thought to contribute to the increase in the trust relationship.

[0064] When the instructor gives advice to a group of trainees at the same time, the trust relationship analysis unit 12 may extract information about the training by referring to the training information DB 53, and may add different degrees of trust relationship construction in the event according to the number of people participating in the training at the same time to the trust relationship index. For example, it may be estimated that the more the number of people participating in the training, the smaller the degree of trust relationship constructed in the event, and the fewer the number of people participating, the greater the degree of trust relationship constructed. More specifically, the increasing trust relationship index may be calculated by dividing the index when one person receives instruction by the number of people participating in the training. For example, it is considered that the smaller the number of people in a conversation, the more seriously the other person's message is taken and the deeper the trust relationship becomes. According to this configuration, it is possible to estimate the trust relationship in accordance with the actual situation.

[0065] The trust relationship analysis unit 12 accumulates a trust relationship index as information on the trust relationship between each instructor and student. The trust relationship analysis unit 12 stores the information on the trust relationship, i.e., the trust relationship index, for example, in a trust relationship index table T8 in the storage unit 50 based on the estimated trust relationship.

[0066] The trust relationship analysis unit 12 may decrease the trust relationship between the lecturer and the students based on the time that has passed since the event that established the trust relationship. In other words, the trust relationship analysis unit 12 may decrease the trust relationship as the time passes since the event. This is because the trust relationship weakens as time passes since the period of interaction. This configuration makes it possible to estimate the trust relationship closer to the actual situation.

[0067] The trust relationship analysis unit 12 may reduce the trust relationship between the instructor and the student depending on the situation of communication between the instructor and the student. For example, when one of the instructors responds negatively to a communication from the other, the trust relationship index between the instructor and the student may be reduced. The degree of reduction in the trust relationship index may be varied depending on the trust relationship index before the event. For example, when the previous trust relationship index is large, the reduced trust relationship index may be reduced. In the case of a relationship in which a relatively sufficient trust relationship has been built, the trust relationship will not be destroyed by a single reaction, and there is a possibility that the person can honestly express their opinion because they trust the other person, and even if there is a misunderstanding, there is a high possibility that they can correct it immediately. In this respect, this configuration makes it possible to estimate the trust relationship in accordance with human sensibilities.

[0068] In addition, the degree of trust to be increased or decreased is not limited to a rule-based estimation process, but may be estimated by any appropriate artificial intelligence technology that references a neural network whose input layer includes, for example, the above-mentioned responses and reactions, the number of calls, the technical level of the trainees, the technical level of the instructor, or the difference in technical levels between the trainees and the instructor, the trust index before the event, the number of training participants, etc.

[0069] The communication processing unit 13 is a processing unit that can execute data transmission and reception processing with the trust relationship estimation device 1 according to a predetermined protocol via a network NW such as the Internet, and is realized by an application, a web browser, or the like. The communication processing unit 13 particularly receives calls from the instructor and the students from the terminal 2. The communication processing unit 13 also receives calls from the instructor from the instructor terminal 2a, and the communication processing unit 13 receives calls from the students from the student terminal 2b. Furthermore, the communication processing unit 13 may transmit the estimated trust relationship between the instructor and the students, or the technical level of the students for each technical field, to the instructor terminal 2a or the student terminal 2b. Furthermore, the communication processing unit 13 may transmit information on the trust relationship between the instructor and the students and the technical level of the students to an appropriate management device connected via the network NW.

[0070] The communication processing unit 13 may also communicate with an appropriate terminal other than the instructor terminal 2a or the student terminal 2b. This terminal is, for example, a terminal used by the student's superior or the student's company, the instructor's superior or the instructor's company, or a person in charge of the training company that plans the training. This terminal may have the same configuration as the instructor terminal 2a or the student terminal 2b, and has an input unit, an output unit, and a communication processing unit.

[0071] The instructor terminal 2a or the student terminal 2b, or the appropriate terminal described above, displays the information received by the communication processing unit 13 on the output unit 22a or the output unit 22b in an appropriate format. FIG. 5 is an example of a screen G10 that displays information about training for a student (hereinafter referred to as student S01) whose identification information is S01. For example, the screen G10 can be displayed on the instructor terminal 2a and the student terminal 2b of the student S01. The screen G10 displays the history of the training attended by the student S01. The history includes, for example, the training content, the attendance schedule, and information on the instructor. The screen G10 also displays, for each technical field, the technical level of the student S01 in that technical field and the trust relationship with the instructor who provided the training in association with each other. In the example of the figure, a scatter diagram is displayed in which the technical level and the trust relationship are shown on two axes as mutually orthogonal indexes. In addition, the figure also displays a scatter diagram for the technical fields G01 and G02. It can be seen that in the technical field G01, the trust relationship with the instructor is well built and the technical level is high, while in the technical field G02, the trust relationship with the instructor is low and the technical level is relatively low.

[0072] The display mode of the history of the training by the trainee may be the same or different on the instructor terminal 2a, the trainee terminal 2b, and other terminals. For example, among the other terminals, a terminal used by a person in charge of the training company may display a list of the trust relationships between a single trainee and multiple instructors. With this configuration, the person in charge of the training company can easily decide the training in which the trainee will participate, taking into account the trust relationship with the instructor. Furthermore, when a system that automatically decides the training mode, such as the training to be held, the instructor, and the participating trainees, is executed on a terminal used by the person in charge of the training company, the system may refer to the trust relationship and decide the training mode so that the trust relationship between the trainee and the instructor in the training to be held will be increased.

[0073] Furthermore, the other terminals may be terminals used by the instructor or the trainee, or the companies to which they belong, or by external companies other than the training company. The external companies are, for example, companies where the trainee hopes to find employment or change jobs. With this configuration, the trainee can show the external companies his or her training attendance history along with the trusting relationship with the instructor. In turn, the trainee can appeal to the external companies where he or she hopes to find employment or change jobs that he or she did not just participate in the training casually, but took the training while building a trusting relationship with the instructor, showing a positive attitude toward the training. In addition, in a terminal used by an external company, the training history of a specific trainee may be displayed only when permission to view the terminal is input from a terminal of the training company that manages the training. Also, the training company's terminal may be able to issue a certificate of the training history for each trainee so that the training history can be viewed by external companies. With this configuration, the training company can present the training history of trainees who have participated in the training managed by the training company to external companies as career information together with the trust relationship of the instructor.

[0074] A sequence diagram of the processing performed by the trust relationship estimation device 1 and each terminal 2 The flow of information between the trust relationship estimation device 1 and each terminal 2 in trust relationship estimation will be described with reference to FIGS. FIG. 6(a) is an example of a mode in which a lecturer and a student transmit messages via their respective terminals 2a and 2b, and is a sequence diagram assuming, for example, an online lecture. In the example shown in FIG. 6, first, the student terminal 2b acquires a question transmitted by the student to the lecturer in the form of audio and video, and transmits it to the trust estimation device 1 (step S101). The student's question audio and video are transmitted to the lecturer terminal 2a via the trust estimation device 1 or directly from the student terminal 2b. Next, the lecturer terminal 2a acquires a response in the form of audio or video, etc., and transmits it to the trust estimation device 1 (step S102). Next, the student terminal 2b acquires a response from the student in the form of audio or video, and transmits it to the trust estimation device 1 (step S103). The student's response is transmitted to the lecturer terminal 2a via the trust estimation device 1 or directly from the student terminal 2b.

[0075] The trust estimation device 1 analyzes the information collected in steps S101 to S103 by the trust analysis unit 12 (step S104). When it detects that the lecturer and the student have alternately sent messages on the same topic at least three times (Y in step S105), it increases the trust index (step S106).

[0076] Fig. 6(b) is a sequence diagram assuming a state in which the instructor and the students transmit data via their respective terminals 2a and 2b. The transmitted data is, for example, an e-mail or chat message sent from the instructor via the instructor terminal 2a, and is, for example, text data, but may also be audio, image, or video data. The data transmitted in steps S111 to S113 may be different types of data. Note that the data transmitted in any of steps S111 to S113 may be a transmission in an online lecture as shown in Fig. 6(a). In the following description, the same processes as those in the sequence diagrams already shown are denoted by the same reference numerals.

[0077] In the example of FIG. 6(b), first, the student terminal 2b accepts the input of the contents of the data and transmits the data to the instructor terminal 2a via the trust relationship estimation device 1 or directly from the student terminal 2b (step S111). Next, the instructor terminal 2a receives the data from the student terminal 2b, and then accepts the input of the contents of the data as a reply, and transmits an email to the student terminal 2b via the trust relationship estimation device 1 or directly from the instructor terminal 2a (step S112). Next, the student terminal 2b receives the data from the instructor terminal 2a, and then accepts the input of the contents of the data as a response, and transmits the data to the instructor terminal 2a via the trust relationship estimation device 1 or directly from the student terminal 2b (step S113). In step S113, instead of or in addition to the configuration of transmitting a specific message to the instructor, when the data from the instructor is displayed or viewed on the student terminal 2b, information indicating that the data has been read may be transmitted to the trust relationship estimation device 1 or the instructor terminal 2a.

[0078] 7(a) is a sequence diagram assuming a mode in which data is input into an input form realized by a web base or a predetermined application. The input form may be, for example, a daily report or a questionnaire. The input form may also be a test administered by an instructor to students. The input form may be realized by a software development platform, and the data to be input in this case may be program code shared between an instructor and students, or a pull request for the program code.

[0079] In the example of FIG. 7(a), first, information on the input form is distributed to the students, and the input form is displayed on the screen of the student terminal 2b (step S121). The input form may be distributed from the trust relationship estimation device 1 or from an appropriate external device. When an input to the input form is accepted as a transmission from the student terminal 2b (step S122), the trust relationship estimation device 1 receives the input information and transmits it to the instructor terminal 2a. The instructor terminal 2a displays the information transmitted from the student terminal 2b and accepts the input as a transmission (step S123). This transmission is, for example, text data including a comment, advice, or guidance on the transmission from the student. The trust relationship estimation device 1 receives the input information and transmits it to the student terminal 2b. The student terminal 2b displays the information transmitted from the instructor terminal 2a and accepts the input as a transmission (step S124).

[0080] The data acquired in steps S122 to S124 may be directly input to the trust relationship estimation device 1 together with the sender's information, instead of being input from the terminal 2. In this case, for example, data written on paper or the like may be read via an appropriate input device such as a scanner or keyboard connected to the trust relationship estimation device 1. In addition, in the above example, all transmissions are made via an input form, but it goes without saying that some may be made in a different format, such as email or chat.

[0081] FIG. 7(b) is a sequence diagram assuming a mode in which calls from a plurality of people are received from one terminal 2, and the trust estimation device 1 separates the callers and analyzes the trust relationship. For example, this mode may be a case in which the voices of the lecturer and the students are widely collected by one terminal 2 provided in a classroom. First, the terminal 2 acquires a plurality of calls from the lecturer or the students (steps S131 to S133), and the trust estimation device 1 receives them. The calls are, for example, voice or video data. Next, the trust estimation device 1 identifies the callers of the calls acquired in steps S131 to S133 using an appropriate voiceprint authentication technique or face identification technique (step S134). The trust estimation device 1 also analyzes the topics that the lecturer and the students are talking about, and analyzes whether they have made calls alternately on the same topic three or more times.

[0082] As described above, the trust estimation device according to the present invention can estimate the strength of the trust relationship built between a lecturer and a student.

[0083] Although the present invention has been described above using the embodiments, the technical scope of the present invention is not limited to the scope described in the above embodiments, and various modifications and changes are possible within the scope of the gist of the present invention. [Explanation of symbols]

[0084] 1. Trust Relationship Estimation Device 11 Input acquisition section 12 Trust Analysis Department 13 Communication processing unit 50 Storage section 2. Terminal 2a Teacher's terminal 2b Participant terminal

Claims

1. A trust relationship estimation device for estimating the trust relationship established between objects, A trust relationship analysis unit detects that, when it detects an event in which transmissions on the same topic have been made alternately at least three times between a first target and a second target different from the first target, it estimates that the trust relationship has been established between the first target and the second target through the transmissions, and stores this information as information relating to the trust relationship in the storage unit. Equipped with, Trust relationship estimation device.

2. The trust relationship analysis unit acquires the transmission by the first target as audio or text data, and analyzes the content of the transmission to estimate that the transmissions between the targets concern the same topic. The reliability relationship estimation device according to claim 1.

3. Further comprising an information acquisition unit that acquires the transmission by the first target as text data, The aforementioned trust relationship analysis unit estimates that the pre-linked and entered transmissions are related to the same topic. The reliability relationship estimation device according to claim 1.

4. The storage unit stores the technical levels of the first and second subjects related to a predetermined technical field. The trust relationship analysis unit adjusts the degree of the trust relationship estimated to have increased between the first and second targets in the event, according to the difference in the technical levels of the first and second targets in the event. The reliability relationship estimation device according to claim 1.

5. The subject has either instructor-side identification information including information that the subject is an instructor, or student-side identification information including information that the subject is a student, The memory unit stores information about the training conducted by the instructor and attended by the participants. The trust relationship analysis unit refers to the memory unit to identify the training in which the event occurred, and determines the degree of trust relationship estimated to have increased between the subject having the instructor's identification information and the subject having the student's identification information in the event, depending on the number of subjects having the student's identification information who are simultaneously participating in the training. The reliability relationship estimation device according to claim 1.

6. The trust relationship analysis unit reduces the degree of the trust relationship between the subjects as time elapses since the event in which the trust relationship between the subjects was established. The reliability relationship estimation device according to claim 1.

7. The information relating to the trust relationship stored in the memory unit is issued for each of the targets. The reliability relationship estimation device according to claim 1.

8. A method for estimating a trust relationship established between objects, A trust relationship analysis step in which, upon detecting an event in which transmissions on the same topic were made alternately at least three times between a first target and a second target different from the first target, it is presumed that a trust relationship was established between the first target and the second target through such transmissions, and this information is stored in the memory unit as information relating to the trust relationship. This is done by a computer. Methods for estimating trust levels.

9. A computer program for estimating trust relationships established between objects, For computers, A trust relationship analysis step in which, upon detecting an event in which transmissions on the same topic were made alternately at least three times between a first target and a second target different from the first target, it is presumed that a trust relationship was established between the first target and the second target through such transmissions, and this information is stored in the memory unit as information relating to the trust relationship. To execute Computer program.

Images