JP7439461B2 - Communication support program, communication support system and communication support method - Google Patents

Description

The present invention relates to a communication support program, a communication support system, and a communication support method, and particularly relates to a communication support program, a communication support system, and a communication support method that support communication using an information terminal.

In recent years, due to work style reforms and other factors, there has been an increase in remote communication using information terminals to send and receive audio and video via communication networks. Even in face-to-face (F2F) communication, problems such as misunderstandings and discrepancies in understanding occur, but problems such as misunderstandings and discrepancies in perception are more likely to occur in remote communication.

Regarding such problems in communication, for example, Patent Document 1 below describes a dialogue support device that mediates communication between a speaker and a listener, which includes a voice recognition unit that recognizes the voice of the speaker, and a voice recognition unit that recognizes the voice of the speaker, An ambiguous expression word list storage unit stores a list of ambiguous expression words that may cause misunderstandings or a shift in consciousness during the process, and the ambiguous expression word list is compared with the speech recognition result. an ambiguous expression detection unit that detects an ambiguous expression phrase including an expression word; an ambiguous expression target detection unit that detects an ambiguous expression target whose interpretation by the listener is ambiguous due to the ambiguous expression phrase from the speech recognition result; a speaker identifying unit that identifies person information and the listener's person information; and a recognition range memory that stores the recognition range of the ambiguous expression for each of the person information for the combination of the ambiguous expression phrase and the ambiguous expression target. and the speaker's cognitive range for the ambiguous expression phrase and the ambiguous expression target related to the speaker's personal information stored in the cognitive range storage, and the listener's personal information stored in the cognitive range storage. a cognitive range estimation unit that calculates the cognitive range of the listener regarding the ambiguous expression phrase and the ambiguous expression target, and compares the cognitive range of the speaker with the cognitive range of the listener to determine the cognitive range of the speaker and the listener. A configuration is disclosed that includes a recognition range difference detection unit for determining a difference.

Japanese Patent Application Publication No. 2008-83753

The comparison between interlocutors in Patent Document 1 is based on ambiguous expressions (slightly, slightly more , change too much, etc.) (cognitive level: length of time, proportion, etc.), not the scope and content of each interlocutor's knowledge (knowledge level: breadth, depth, density, etc. of knowledge). Therefore, even if the technique of Patent Document 1 is used, it is not possible to solve problems such as misunderstandings and discrepancies in understanding that occur due to differences in knowledge levels of individuals regarding communication topics.

The present invention has been made in view of the above problems, and its main purpose is to provide a communication support program, a communication support system, and a communication support method that can avoid misunderstandings and misunderstandings in communication using information terminals. Our goal is to provide the following.

One aspect of the present invention is a communication support program that operates on a device that supports communication between interlocutors, wherein a control unit of the device includes a first acquisition process for acquiring keywords in the communication, and a first acquisition process for acquiring keywords in the communication. a second acquisition process of acquiring, for each interlocutor, a knowledge network representing connections between words related to the keyword as the person's knowledge regarding the keyword; and information indicating differences between the knowledge networks of each interlocutor. The present invention is characterized by causing output processing to output.

One aspect of the present invention is a communication support system, characterized in that a plurality of information terminals with which each individual communicates and the device in which the communication support program is installed are connected via a communication network. do.

One aspect of the present invention is a communication support system, characterized in that a plurality of information terminals with which individuals communicate are connected via a communication network, and the communication support program is installed in each of the information terminals. shall be.

One aspect of the present invention is a communication support method in a device that supports communication between interlocutors, including a first acquisition process for acquiring keywords in the communication, and knowledge regarding the keywords of each interlocutor. a second acquisition process of acquiring a knowledge network representing a connection between words for the keyword for each of the interlocutors; and an output process of outputting information indicating a difference between the knowledge networks of each of the interlocutors; It is characterized by carrying out.

According to the communication support program, remote communication support system, and communication support method of the present invention, misunderstandings and misunderstandings in communication using an information terminal can be avoided.

The reason for this is that the device supports communication between interlocutors, and the first acquisition process acquires keywords in communication, and the knowledge network representing the connections between words for keywords is used as knowledge about the keywords of each interlocutor. This is because the second acquisition process is performed for each interlocutor, and the output process is executed to output information indicating the differences between the knowledge networks of each interlocutor.

1 is a schematic diagram showing an example of the configuration of a communication support system according to an embodiment of the present invention. FIG. 2 is a schematic diagram showing another example of the configuration of a communication support system according to an embodiment of the present invention. FIG. 1 is a block diagram showing the configuration of an information terminal according to an embodiment of the present invention. FIG. 1 is a block diagram showing the configuration of a communication support device according to an embodiment of the present invention. FIG. 2 is a schematic diagram showing the general operation of a communication support method according to an embodiment of the present invention. It is an example of the ID management table based on one Example of this invention. It is an image diagram showing knowledge regarding keywords. It is an image diagram showing differences in knowledge regarding keywords. FIG. 2 is a diagram illustrating a keyword acquisition method according to an embodiment of the present invention. FIG. 3 is a flowchart showing the operation of the communication support device according to an embodiment of the present invention. FIG. 3 is a flowchart showing the operation of the communication support device according to an embodiment of the present invention.

As described in the Background Art section, problems such as misunderstandings and discrepancies in recognition are likely to occur in remote communication that uses information terminals to transmit and receive audio and video via communication networks. In response to such communication problems, Patent Document 1 detects ambiguous expressions and their targets, identifies the cognitive range of the speaker/listener in these from a cognitive range storage unit, and notifies the user if the difference exceeds a certain level. That's what I do. However, this technique specializes in misunderstandings caused by ambiguous expressions, and does not take into account differences in individuals' knowledge levels regarding communication topics. Therefore, problems such as misunderstandings and discrepancies in understanding that occur due to differences in the knowledge level of each individual cannot be avoided.

Therefore, in one embodiment of the present invention, each individual's knowledge level with respect to communication keywords (topics, topics, technical terms, topic-related terms, etc.) is estimated, and based on the difference in knowledge level, misunderstandings and perceptions in communication are The possibility of a discrepancy occurring is determined and notified.

Specifically, in a device that supports communication between interlocutors, the first acquisition process acquires keywords in communication, and the knowledge network representing the connections between words corresponding to the keywords is acquired as knowledge regarding the keywords of each interlocutor. A second acquisition process for acquiring each interlocutor and an output process for outputting information indicating the differences between the knowledge networks of each interlocutor are executed. Furthermore, a process for identifying interlocutors that identifies the interlocutors who are communicating, a knowledge difference extraction process that compares the knowledge networks of each interlocutor to extract differences in knowledge regarding keywords, and a process for identifying differences in knowledge regarding keywords. Accordingly, a notification determination process for determining the necessity of notification is executed, and in the output process, when it is determined that there is a need for notification, it is notified that there is a difference in knowledge regarding the keyword.

In this way, if there is a certain level of knowledge difference regarding a communication keyword, by informing the communication interlocutors of this fact, they can communicate with each other while recognizing that there is a knowledge difference. , it is possible to prevent the occurrence of situations in which discussions do not fit together due to unexpected misunderstandings or discrepancies in understanding.

In order to describe the embodiment of the present invention described above in more detail, a communication support program, a communication support system, and a communication support method according to an embodiment of the present invention will be described with reference to FIGS. 1 to 11. 1 and 2 are schematic diagrams showing the configuration of the communication support system of this embodiment, FIG. 3 is a block diagram showing the configuration of the information terminal, and FIG. 4 is a block diagram showing the configuration of the communication support device. be. Further, FIG. 5 is a schematic diagram showing an outline of the communication support method of this embodiment, and FIG. 6 is an example of an ID management table of this embodiment. Further, FIG. 7 is an image diagram showing knowledge regarding keywords, and FIG. 8 is an image diagram showing differences in knowledge regarding keywords. Moreover, FIG. 9 is a diagram explaining the keyword estimation method of this embodiment, and FIGS. 10 and 11 are flowcharts showing the operation of the communication support device of this embodiment.

As shown in FIG. 1, the communication support system 10 of this embodiment includes a plurality of information terminals 20 and a communication support device 30 used by individuals who communicate, and these are connected to Ethernet (registered trademark), token ring, They are connected via a communication network 40 such as a LAN (Local Area Network) or a WAN (Wide Area Network) defined by standards such as FDDI (Fiber-Distributed Data Interface).

In FIG. 1, the information terminal 20 and the communication support device 30 are separate devices, but as shown in FIG. 2, the communication support device 30 may be built in each information terminal 20. Each device will be described in detail below based on the system configuration shown in FIG.

[Information terminal]
The information terminal 20 is a personal computer, tablet, smartphone, or the like that communicates through audio and video. As shown in FIG. 3A, this information terminal 20 includes a control section 21, a storage section 22, a network I/F section 23, a video input section 24, an audio input/output section 25, a display section 26, an operation section 27, etc. Consists of.

The control unit 21 is composed of a CPU (Central Processing Unit) constituted by a microprocessor or the like, and memory such as a ROM (Read Only Memory) or a RAM (Random Access Memory). The overall operation of the information terminal 20 is controlled by expanding the control program into the RAM and executing it.

Further, as shown in FIG. 3(b), the control unit 21 executes an interaction application 21a and an information collection application 21b.

The dialogue application 21a realizes two-way communication between clients via an IP (Internet Protocol) telephone, Skype (registered trademark), and a server that realizes two-way communication using P2P (peer-to-peer) technology. Applications such as electronic conference systems.

The information collection application 21b acquires information regarding the knowledge of the individual who uses the information terminal 20. For example, when installing the information collection application 21b on each information terminal 20, a personal ID is input. The information collection application 21b is operated as a resident application, and documents displayed on the display unit 26, information sent and received via e-mail, SNS (Social Networking Service), IP telephone, etc., web information viewed using a web browser, Obtain information such as audio and minutes of electronic meetings. Specifically, the information collection application 21b acquires the information displayed on the screen, acquires text information using well-known OCR (Optical Character Recognition), and associates the acquired text information with a personal ID. It is stored in the storage unit 22 or notified to the communication support device 30. The information collection application 21b also acquires voice information, converts it into text information using a known voice recognition technology, associates the converted text information with the personal ID, and stores it in the storage unit 22, or the communication support device 30 will be notified.

The storage unit 22 is composed of an HDD (Hard Disk Drive), an SSD (Solid State Drive), etc., and contains programs for the CPU to control each unit (including the dialogue application 21a and the information collection application 21b), and the processing of the own device. It stores information related to functions, information collected by the information collection application 21b, and the like.

The network I/F unit 23 is composed of a NIC (Network Interface Card), a modem, etc., connects the information terminal 20 to the communication network 40, enables two-way video and audio communication between the information terminals 20, and , information collected by the information collection application 21b is associated with a personal ID and transmitted to the communication support device 30, and communication information between the information terminals 20 is transmitted to the communication support device 30.

The video input unit 24 is configured with a CCD (Charge Coupled Devices) camera or the like, and acquires a video of an individual operating the information terminal 20. The audio input/output unit 25 is composed of a microphone, a speaker, and the like, and acquires the audio of an individual operating the information terminal 20 and outputs the audio of an individual operating another information terminal 20.

The display unit 26 includes a liquid crystal display (LCD), an organic EL (electroluminescence) display, and the like, and displays a screen of the interactive application 21a. The operation unit 27 includes a mouse, a keyboard, etc., and enables various operations on the screen of the dialog application 21a.

[Communication support device]
The communication support device 30 is a computer device or the like that supports communication between the information terminals 20. As shown in FIG. 4(a), the communication support device 30 includes a control section 31, a storage section 32, a network I/F section 33, a notification section 34, and the like.

The control unit 31 is composed of a CPU constituted by a microprocessor or the like and a memory such as a ROM or a RAM, and the CPU executes a control program (including the communication support program of this embodiment) stored in the ROM or the storage unit 32. The operation of the entire communication support device 30 is controlled by expanding and executing the program in the RAM.

Further, as shown in FIG. 4(b), the control section 31 includes a knowledge information extraction section 31a, an interlocutor identification section 31b, a speech recognition section 31c, a keyword acquisition section 31d, a keyword knowledge extraction section 31e, and a knowledge difference extraction section. 31f, the notification determination section 31g, etc.

The knowledge information extraction unit 31a acquires information associated with a personal ID from the information terminal 20, and performs morphological analysis to break down sentences and phrases included in the acquired information into words, which are the minimum units that have meaning. A co-occurrence degree analysis is performed on the decomposed words to generate a network model (for example, a co-occurrence network) representing the connections between words. The generated co-occurrence network is then registered in a knowledge network (referred to as a personal knowledge DB in this embodiment) in association with a personal ID. Further, when acquiring new information, the knowledge information extraction unit 31a identifies the target personal knowledge DB based on the personal ID and merges the information. In this embodiment, it is assumed that network information is registered in the personal knowledge DB.

Upon receiving the communication information from the information terminal 20, the interlocutor specifying unit 31b refers to the personal identification information in the ID management table stored in the storage unit 32, identifies the interlocutor who is communicating, and identifies the interlocutor. Obtain the person's personal ID. For example, when voice information is received from the information terminal 20, the interlocutor is identified by comparing it with the voiceprint information in the ID management table, and the individual ID of the interlocutor is obtained.

The voice recognition unit 31c is provided as necessary, and when the communication information is voice information, it converts the voice information into text information using a known voice recognition technique.

The keyword acquisition unit 31d acquires keywords in communication (things that are the content of communication, including topics, agendas, technical terms, and terms related to topics) (this process is referred to as a first acquisition process). For example, communication keywords are estimated from the content of the communication. Specifically, we perform morphological analysis on communication information (text information), generate a network model (e.g., co-occurrence network), identify the central node of the co-occurrence network using proximity centrality, etc. The node extracted as the center is estimated as a "keyword". Note that here, a co-occurrence network is used as a network model and keywords are estimated using closeness centrality, but the keyword estimation method is not limited to the above method. Further, in this embodiment, a case will be described in which a co-occurrence network is generated and a keyword is estimated, but if a keyword is presented during communication, the keyword may be acquired.

The keyword knowledge extraction unit 31e acquires a knowledge network representing the connections between words with respect to the keywords for each interlocutor as knowledge regarding the keyword of each interlocutor (this process is referred to as a second acquisition process). For example, using the personal ID specified by the interlocutor identification unit 31b and the keyword acquired (estimated) by the keyword acquisition unit 31d, the personal knowledge DB stored in the storage unit 32 is referred to, and the keyword of each interlocutor is knowledge (referred to as keyword knowledge) is extracted as a network representing the connections between words. The method for extracting this keyword knowledge will be described later.

The knowledge difference extraction unit 31f compares the networks of each interlocutor extracted by the keyword knowledge extraction unit 31e, and extracts differences in the keyword knowledge of each interlocutor. For example, the nodes existing in the keyword knowledge of each interlocutor are compared, and the difference rate (knowledge difference) is extracted.

The notification determination unit 31g determines whether or not notification is necessary, depending on the difference in keyword knowledge of each interlocutor extracted by the knowledge difference extraction unit 31f. For example, if the difference rate extracted by the knowledge difference extraction unit 31f exceeds a predetermined threshold, it is determined that there is a need for notification because misunderstandings or discrepancies in communication are likely to occur, and it is determined that there is a difference in knowledge regarding the keyword. The notifying unit 34 is made to notify that.

The knowledge information extraction unit 31a, interlocutor identification unit 31b, voice recognition unit 31c, keyword acquisition unit 31d, keyword knowledge extraction unit 31e, knowledge difference extraction unit 31f, and notification determination unit 31g may be configured as hardware, The control unit 31 includes a knowledge information extraction unit 31a, an interlocutor identification unit 31b, a speech recognition unit 31c, a keyword acquisition unit 31d, a keyword knowledge extraction unit 31e, a knowledge difference extraction unit 31f, and a notification determination unit 31g (in particular, an interlocutor identification unit 31b, keyword acquisition unit 31d, keyword knowledge extraction unit 31e, knowledge difference extraction unit 31f, and notification determination unit 31g), and the CPU is configured to execute the communication support program. good.

The storage unit 32 is composed of an HDD, an SSD, etc., and stores programs for the CPU to control each unit, information regarding the processing functions of the own device, a personal knowledge DB, an ID management table shown in FIG. 6, and the like.

The network I/F unit 33 is configured with a NIC, a modem, etc., and connects the communication support device 30 to the communication network 40, and acquires information and communication information collected by the information collection application 21b from the information terminal 20, and as necessary. Accordingly, each information terminal 20 is notified of broadcast information.

The notification unit 34 is a display unit, an output unit, or the like, and outputs information indicating the differences between the knowledge networks of each interlocutor (this processing is referred to as output processing). For example, the notification unit 34 reports that there is a difference in knowledge regarding the keyword, according to the determination result of the notification determination unit 31g. Further, the notification unit 34 displays the image diagram of the co-occurrence network shown in FIGS. 7 and 8, or displays the knowledge difference rate, according to instructions from the keyword knowledge extraction unit 31e and the knowledge difference extraction unit 31f.

Note that FIGS. 1 to 5 are examples of the communication support system 10 of this embodiment, and the configuration and control of each device can be changed as appropriate.

For example, in this embodiment, the personal knowledge DB is stored in the storage unit 32 of the communication support device 30, but the personal knowledge DB is stored in a server connected via the communication network 40, and the keyword knowledge extraction unit 31e The personal knowledge DB may also be referenced by accessing the above server or the like.

In addition, in FIG. 4, the notification unit 34 is provided in the communication support device 30 to notify that there is a difference in knowledge regarding keywords, but notification information is notified to the information terminal 20 according to the determination result of the notification determination unit 31g, and information The audio input/output unit 25 and display unit 26 of the terminal 20 may notify that there is a difference in knowledge regarding the keyword. In this case, all interlocutors may be informed that there are differences in knowledge regarding keywords, or some interlocutors (for example, interlocutors with extensive knowledge of keywords) may be informed that there are differences in knowledge regarding keywords. You may notify that something is going on.

Further, in FIG. 4, the communication support device 30 has been described as a separate device from the information terminal 20, but as shown in FIG. 2, the communication support device 30 may be included in the information terminal 20, in which case, , the control unit 21 of the information terminal 20 in FIG. 31g.

Processing when communication information is input to the communication support system 10 will be described below with reference to FIGS. 5 to 8. Here, we will explain using voice communication as an example. It is assumed that the personal knowledge DB and the ID management table are created in advance and stored in the storage unit 32 or the like.

(1) Identification of interlocutor The interlocutor identification unit 31b acquires voice information as communication information. Then, by comparing it with the individual identification information (voiceprint information) in the ID management table shown in FIG. 6, the individual corresponding to the acquired voice information is identified and the individual ID (here, u0001, u0003) is acquired. , notifies the keyword knowledge extraction unit 31e of the identified personal ID.

(2) Estimation of keywords The speech recognition unit 31c converts the utterance content (speech information) into text information, and the keyword acquisition unit 31d generates a co-occurrence network of the transcribed utterance content. Then, the central node of the co-occurrence network is extracted using proximity centrality, the extracted node is estimated as a keyword, and the estimated keyword is notified to the keyword knowledge extraction unit 31e. Here, explanation will be given assuming that "hypertension" is estimated as a keyword.

(3) Extraction of keyword knowledge The keyword knowledge extraction unit 31e, which has been notified of the personal ID and keyword, refers to the personal knowledge DB and extracts the keyword knowledge for the personal ID (u0001, u0003 in this case). Keyword knowledge of each interlocutor is also network information representing connections between words. Then, using the keyword "hypertension" as the root (center), information (keyword group) connected to it is extracted. At this time, it is preferable to visualize the keyword knowledge so that it can be intuitively understood. For example, an image diagram representing a network as shown in FIG. 7 can be displayed.

(4) Extraction of knowledge differences (node match rate)
The nodes existing in the keyword knowledge of each interlocutor extracted in the process (3) above are compared. For example, assume that there are 120 nodes of u0001 and 90 nodes of u0003, and that 60 nodes match on both sides. The difference rate at this time can be expressed as 1-common part/union,
Difference rate = 1-60/(120+90-60)=0.6
becomes. In this case, it is preferable to visualize the differences in keyword knowledge as network differences so that they can be intuitively understood. For example, as shown in FIG. 8(a), images of the keyword knowledge of each interlocutor are displayed for comparison. It is also possible to display the difference rate along with the keyword knowledge of each interlocutor, as shown in Figure 8(b), and to highlight the differences (see the dotted hatched area in the figure). . Note that other methods for extracting knowledge differences will be described later.

(5) Determination and Notification of Notification If the difference rate extracted by the knowledge difference extraction unit 31f exceeds a predetermined threshold, the notification determination unit 31g determines that notification is necessary because misunderstandings in communication and recognition discrepancies are likely to occur. judge. For example, if "0.5" is set as the difference rate threshold, in the above example, the threshold is exceeded, so it is determined that notification is necessary. If it is determined that notification is necessary, the notification unit 34 notifies the user that there is a difference in keyword knowledge using audio information or visual information. Alternatively, notification information may be transmitted to the information terminal 20 to notify that there is a difference in keyword knowledge using audio information or visual information from the audio input/output unit 25 or display unit 26 of the information terminal 20.

Note that in (4) above, the knowledge difference is extracted using the matching rate of nodes, but the knowledge difference can also be extracted using the distance between nodes. Specifically, each node existing in the keyword knowledge of each interlocutor extracted in the process (3) is converted into a vector value. As a method for converting into a vector value, a method such as Word2Vec can be used, for example. Next, the average value of the vector values of each node is calculated, and the average value is used as the knowledge vector of each interlocutor.
u0001: (V1+V2...+V120)/120=u01V
u0003: (V1+V2...+V90)/90=u03V
Then, the similarity of each vector value is determined using cosine similarity (distance).
When cos (u01V, u03V) = 0.73, since 0.73 is the degree of similarity, the difference rate is 1-0.73 = 0.27, and the knowledge difference can be extracted based on this difference rate.

Furthermore, knowledge differences can also be extracted using the way nodes are connected. Specifically, for each interlocutor, the number of routes from the root to all nodes is calculated. Here, u0001: 100 and u0003: 120, and the number of routes that match between the interlocutors is calculated. The difference rate at this time can be expressed as 1-common part/union. For example, if 30 routes match,
Difference rate=1-30/(100+120-30)=0.84, and knowledge differences can be extracted based on this difference rate.

Furthermore, knowledge differences can be extracted for nodes having a strength of association greater than a predetermined value. Specifically, the keyword knowledge of each interlocutor extracted in the process (3) is a network with the keyword "hypertension" as the root. Knowledge that is far from the keyword is judged to be unimportant; for example, information (nodes) from the root to the fifth layer is used, and the layers after that are ignored. Then, knowledge differences (difference rates) can be extracted for nodes up to the fifth layer based on the node match rate, distance between nodes, and how nodes are connected.

Furthermore, the processes (1) to (5) can be dynamically changed in accordance with changes in the content of communication. Dialogue is progressing from moment to moment, and new communication information is input from time to time. Therefore, the keyword acquisition unit 31d performs estimation every time it acquires a certain amount of communication information. For example, as shown in FIG. 9, the keyword acquisition unit 31d periodically monitors communication information, performs estimation again when new communication information of 300 characters or more is input, or performs estimation again after a predetermined period of time has elapsed. Then, estimate again. Furthermore, a certain amount of the latest information can be used for keyword estimation. For example, if the upper limit is 1000 characters, keyword estimation is performed using communication information up to 1000 characters. If the keyword changes, the keyword knowledge extraction in (3), the knowledge difference extraction in (4), and the notification determination and notification in (5) are performed again for the re-estimated keyword.

Hereinafter, specific processing of the communication support device 30 will be explained. The CPU executes each step shown in the flowcharts of FIGS. 10 and 11 by loading the communication support program stored in the ROM or storage unit 32 into the RAM and executing it.

First, the control unit 31 (interlocutor identifying unit 31b) identifies the interlocutor from communication information and acquires the individual ID (S101). If the interlocutor cannot be identified (No in S102), it is notified that the interlocutor is not an individual in the personal knowledge DB (S103), and the series of processes ends.

On the other hand, if the interlocutor can be identified (Yes in S102), the control unit 31 (speech recognition unit 31c) converts the voice information into text information, and the control unit 31 (keyword acquisition unit 31d) converts the voice information into text information as necessary. estimates keywords from the conversation content (S104). Next, the control unit 31 (keyword knowledge extraction unit 31e) accesses the personal knowledge DB of each interlocutor using the acquired personal ID (S105), and uses the estimated keyword to access the personal knowledge DB of each interlocutor. Keyword knowledge is extracted from among (S106). At this time, the extracted keyword knowledge of each interlocutor can be visualized in an image diagram as shown in FIG.

After executing the process of S106 for each interlocutor, the control unit 31 (knowledge difference extraction unit 31f) extracts difference information (difference rate) of keyword knowledge for each interlocutor (S107). At this time, the difference information (difference rate) of the extracted keyword knowledge can be visualized in an image diagram as shown in FIG.

After that, the control unit 31 (notification determination unit 31g) determines the necessity of notification from the extracted keyword knowledge difference information (difference rate) (S108). If it is determined that notification is not necessary, the process returns to S101 and the communication information is monitored, and if it is determined that notification is necessary, the notification unit 34 notifies that there is a difference in keyword knowledge between the interlocutors ( S109), the series of processing ends.

FIG. 10 shows the process when communication is being carried out regarding a certain keyword, but the communication changes from moment to moment, and the keyword also changes accordingly. In that case, as shown in FIG. 11, the control unit 31 (keyword acquisition unit 31d) estimates the keyword periodically (for example, every certain number of characters or every certain period of time) (S201). Next, the control unit 31 (keyword acquisition unit 31d) determines whether there is a change in the keyword (S202), and if there is no change (No in S202), returns to S201 to estimate the keyword, and If there is a keyword (Yes in S202), the process shown in FIG. 10 is performed on the new keyword (S203).

In this way, if there is a certain level of difference in knowledge regarding a keyword, by informing the interlocutors of this fact, they can communicate with each other while being aware of the difference in knowledge. It is possible to prevent the occurrence of situations where discussions do not fit together due to misunderstandings or discrepancies in understanding.

It should be noted that the present invention is not limited to the description of the above embodiments, and the configuration and control thereof can be modified as appropriate without departing from the spirit of the present invention.

For example, in the above embodiment, the personal knowledge DB is stored in the storage unit 32 of the communication support device 30, but the personal knowledge DB is stored in a server on a network, and the communication support device 30 stores the personal knowledge of the server. A configuration may also be adopted in which the keyword knowledge of each interlocutor is acquired by accessing the DB.

INDUSTRIAL APPLICATION This invention can be utilized for the communication support program which supports communication using an information terminal, the recording medium which recorded the said communication support program, a communication support system, and the communication support method.

10 Communication support system 20 Information terminal 21 Control unit 21a Dialogue application 21b Information collection application 22 Storage unit 23 Network I/F unit 24 Video input unit 25 Audio input/output unit 26 Display unit 27 Operation unit 30 Communication support device 31 Control unit 31a Knowledge Information extraction unit 31b Interlocutor identification unit 31c Speech recognition unit 31d Keyword acquisition unit 31e Keyword knowledge extraction unit 31f Knowledge difference extraction unit 31g Notification determination unit 32 Storage unit 33 Network I/F unit 34 Notification unit 40 Communication network

Claims (28)

A communication support program that operates on a device that supports communication between interlocutors,
In the control section of the device,
interlocutor identification processing for identifying the interlocutor who is performing the communication;
a first acquisition process for acquiring keywords in the communication;
a second acquisition process of acquiring for each interlocutor a knowledge network representing connections between words with respect to the keyword as knowledge regarding the keyword of each interlocutor;
knowledge difference extraction processing that compares the knowledge networks of each of the interlocutors and extracts differences in knowledge regarding the keyword;
executing an output process of outputting information for indicating differences between the extracted knowledge networks of the interlocutors;
A communication support program characterized by: In the first acquisition process, a keyword of the communication is estimated from the content of the communication.
The communication support program according to claim 1, characterized in that: In the first acquisition process, a predetermined amount of audio information and/or text information is acquired, or after a predetermined time has elapsed, the keyword is re-estimated.
The communication support program according to claim 2, characterized in that: The control unit of the device further includes :
Executing a notification determination process to determine the necessity of notification according to the difference in knowledge regarding the keyword,
In the output process, if it is determined that there is a need for notification, notification is made that there is a difference in knowledge regarding the keyword;
The communication support program according to any one of claims 1 to 3, characterized in that: The control unit of the device further includes:
If the communication is voice, executing voice recognition processing to convert voice information into text information;
5. The communication support program according to claim 4. In the knowledge difference extraction process, a difference in knowledge regarding the keyword is extracted based on a matching rate of nodes related to the keyword in the knowledge network.
The communication support program according to any one of claims 1 to 5, characterized in that: In the knowledge difference extraction process, a difference in knowledge regarding the keyword is extracted based on a distance between nodes related to the keyword in the knowledge network.
The communication support program according to any one of claims 1 to 5, characterized in that: The knowledge difference extraction process extracts knowledge differences regarding the keyword based on how nodes related to the keyword in the knowledge network are connected.
The communication support program according to any one of claims 1 to 5, characterized in that: In the knowledge difference extraction process, a difference in knowledge regarding the keyword is extracted from nodes that have a strength of association with the keyword of a predetermined value or higher.
The communication support program according to any one of claims 6 to 8, characterized in that: In the knowledge difference extraction process, the difference in knowledge regarding the extracted keyword is visualized as a difference in the knowledge network.
The communication support program according to any one of claims 1 to 9 , characterized in that: The device is a device that can refer to the knowledge network and acquire communication audio information and/or text information from an information terminal,
The control unit of the device further includes:
Text information viewed by the individual and/or voice information pronounced by the individual is acquired from the information terminal, knowledge information of the individual is extracted from the text information and/or voice information, and the extracted knowledge information and the voice information are extracted from the text information and/or the voice information. Execute knowledge information extraction processing that associates personal identification information with the knowledge network and registers it in the knowledge network;
The communication support program according to any one of claims 1 to 10. The knowledge network is a personal knowledge DB that stores personal knowledge for each individual.
The communication support program according to any one of claims 1 to 11. the keyword is the topic of the communication;
The communication support program according to any one of claims 1 to 12, characterized in that: A plurality of information terminals with which each individual communicates and the device in which the communication support program according to any one of claims 1 to 13 is installed are connected via a communication network.
A communication support system characterized by: A plurality of information terminals with which each individual communicates are connected via a communication network, and each of the information terminals is installed with the communication support program according to any one of claims 1 to 13.
A communication support system characterized by: A communication support method in a device for supporting communication between interlocutors, the method comprising:
interlocutor identification processing for identifying the interlocutor who is performing the communication;
a first acquisition process of acquiring keywords in the communication;
a second acquisition process that acquires for each interlocutor a knowledge network representing connections between words with respect to the keyword as knowledge regarding the keyword of each interlocutor;
knowledge difference extraction processing that compares the knowledge networks of each of the interlocutors and extracts differences in knowledge regarding the keyword;
outputting information for indicating differences between the extracted knowledge networks of the interlocutors;
A communication support method characterized by: In the first acquisition process, a keyword of the communication is estimated from the content of the communication.
17. The communication support method according to claim 16. In the first acquisition process, a predetermined amount of audio information and/or text information is acquired, or after a predetermined time has elapsed, the keyword is re-estimated.
The communication support method according to claim 17, characterized in that: further executing a notification determination process to determine the necessity of notification according to the difference in knowledge regarding the keyword;
In the output process, if it is determined that there is a need for notification, notification is made that there is a difference in knowledge regarding the keyword;
The communication support method according to any one of claims 16 to 18. If the communication is voice information, further performing voice recognition processing to convert the voice information into text information;
20. The communication support method according to claim 19. In the knowledge difference extraction process, a difference in knowledge regarding the keyword is extracted based on a matching rate of nodes related to the keyword in the knowledge network.
The communication support method according to any one of claims 16 to 20 . In the knowledge difference extraction process, a difference in knowledge regarding the keyword is extracted based on a distance between nodes related to the keyword in the knowledge network.
The communication support method according to any one of claims 16 to 20 . The knowledge difference extraction process extracts knowledge differences regarding the keyword based on how nodes related to the keyword in the knowledge network are connected.
The communication support method according to any one of claims 16 to 20 . In the knowledge difference extraction process, a difference in knowledge regarding the keyword is extracted from nodes that have a strength of association with the keyword of a predetermined value or higher.
24. The communication support method according to any one of claims 21 to 23. In the knowledge difference extraction process, the difference in knowledge regarding the extracted keyword is visualized as a difference in the knowledge network.
The communication support method according to any one of claims 16 to 24. The device is a device that can refer to the knowledge network and acquire communication audio information and/or text information from an information terminal,
Text information viewed by the individual and/or voice information pronounced by the individual is acquired from the information terminal, knowledge information of the individual is extracted from the text information and/or voice information, and the extracted knowledge information and the voice information are extracted from the text information and/or the voice information. further executing a knowledge information extraction process of associating the information with personal identification information and registering it in the knowledge network;
The communication support method according to any one of claims 16 to 25. The knowledge network is a personal knowledge DB that stores personal knowledge for each individual.
27. The communication support method according to any one of claims 16 to 26. the keyword is the topic of the communication;
28. The communication support method according to any one of claims 16 to 27.

Images