JP7623968B2 - System and method for supporting web conferencing - Google Patents

Description

The present invention generally relates to computer technology for supporting web conferencing.

Recently, web conferences, where participants meet over the internet without meeting face to face, have become increasingly popular in a variety of business fields.

One specific example of a business sector in which web conferencing is becoming more common is business negotiations conducted online as part of sales activities for products or services.

In addition, various systems have been proposed for supporting sales activities for products and services (for example, Patent Document 1).

JP 2003-157354 A

In the case of web conferences, by recording the conference, various digital data related to the speaker's emotions, speech content, etc. can be easily extracted and generated from the recorded conference data. Therefore, it is conceivable that such technology can be used to support web conferences for online sales activities and other purposes.

The system includes a storage device and a processor connected to the storage device. The storage device stores recorded data for each web conference. For each piece of recorded data, the type of topic in the web conference represented by the recorded data is associated in the storage device. For each piece of recorded data, the processor calculates an attention level, which is an evaluation value based on the facial expression, movement, and speech content of at least one participant recorded in the recorded data. When the processor receives a piece of recorded data designated by a user, the processor retrieves from the storage device recorded data that has the same type of topic as the designated recorded data and has a difference in attention level with the designated recorded data within a predetermined range, and displays information related to the retrieved recorded data.

The present invention provides web conference support that contributes to achieving the objectives of web conferences.

1 shows an overview of an embodiment. 1 illustrates an example of the configuration of a web conference support system according to an embodiment. 13 shows an example of the configuration of a video management table. 13 shows an example of the configuration of a video analysis result table. 13 shows an example of the configuration of an emotion type table. 13 shows an example of the configuration of a conference information management table. 13 shows an example of the configuration of a customer management table. An example of a customer type is shown below. 1 shows an example of the overall flow of processing performed in an embodiment. 13 shows an example of a flow of a video analysis process. 13 shows an example of the flow of a meeting information extraction process. 13 shows an example of the flow of an attention level calculation process. 13 shows an example of the flow of a reference information extraction process. 13 shows an example of the flow of a difference point calculation process. 13 shows an example of a reference information display screen. 1 shows an example of a timeline display screen. 1 shows an example of a screen displaying important/bad words. An example of an important document display screen is shown. 13 shows an example of the flow of a recommendation target extraction process.

In the following description, an "interface unit" may refer to one or more interface devices. The one or more interface devices may be at least one of the following:
One or more I/O (Input/Output) interface devices. The I/O (Input/Output) interface devices are interface devices to at least one of the I/O devices and a remote display computer. The I/O interface device to the display computer may be a communications interface device. The at least one I/O device may be a user interface device, for example, either an input device such as a keyboard and a pointing device, or an output device such as a display device.
One or more communication interface devices. The one or more communication interface devices may be one or more homogeneous communication interface devices (e.g., one or more NICs (Network Interface Cards)) or two or more heterogeneous communication interface devices (e.g., a NIC and an HBA (Host Bus Adapter)).

In the following description, "memory" refers to one or more memory devices, which are an example of one or more storage devices, and may typically be a primary storage device. At least one memory device in the memory may be a volatile memory device or a non-volatile memory device.

In the following description, a "persistent storage device" may be one or more persistent storage devices, which are an example of one or more storage devices. A persistent storage device may typically be a non-volatile storage device (e.g., an auxiliary storage device), and more specifically, may be, for example, a hard disk drive (HDD), a solid state drive (SSD), a non-volatile memory express (NVME) drive, or a storage class memory (SCM).

In the following description, "storage device" may refer to at least one memory, including memory and persistent storage device.

In the following description, a "processor" may be one or more processor devices. Typically, the at least one processor device may be a microprocessor device such as a CPU (Central Processing Unit), but may also be other types of processor devices such as a GPU (Graphics Processing Unit). The at least one processor device may be a single-core or multi-core. The at least one processor device may be a processor core. The at least one processor device may also be a broader processor device such as a circuit that is a collection of gate arrays written in a hardware description language that performs part or all of the processing (e.g., an FPGA (Field-Programmable Gate Array), a CPLD (Complex Programmable Logic Device), or an ASIC (Application Specific Integrated Circuit)).

In the following description, functions are sometimes described using the expression "yyy unit", but the functions may be realized by one or more computer programs being executed by a processor, or by one or more hardware circuits (e.g., FPGAs or ASICs), or by a combination thereof. When a function is realized by a program being executed by a processor, the function may be at least a part of the processor, since the specified processing is performed using a storage device and/or an interface device, etc., as appropriate. Processing described with a function as the subject may be processing performed by a processor or a device having the processor. A program may be installed from a program source. The program source may be, for example, a program distribution computer or a computer-readable recording medium (e.g., a non-transitory recording medium). The description of each function is an example, and multiple functions may be combined into one function, or one function may be divided into multiple functions.

In addition, in the following explanation, the processing may be explained with a "program" as the subject, but the processing explained with a program as the subject may also be processing performed by a processor or a device having that processor. Furthermore, two or more programs may be realized as one program, and one program may be realized as two or more programs.

In the following explanation, information that gives an output for an input may be described using expressions such as "xxx table", but the information may be a table of any structure, or may be a neural network that generates an output for an input, or a learning model such as a genetic algorithm or random forest. Therefore, a "xxx table" can be called "xxx information". In the following explanation, the structure of each table is an example, and one table may be divided into two or more tables, or all or part of two or more tables may be one table.

In the following description, the "web conference support system" may be a system composed of one or more physical computers, or may be a system (e.g., a cloud computing system) implemented on a group of physical computing resources (e.g., a cloud platform). When the web conference support system "displays" display information, it may mean displaying the display information on a display device possessed by the computer, or the computer may send the display information to a display computer (in the latter case, the display information is displayed by the display computer).
The following description of the embodiment will be given taking as an example a business negotiation for concluding a sales contract for a product or service, which is conducted online between one customer and one sales representative using a well-known web conferencing tool.

In this case, the business negotiations will usually proceed step by step through multiple web conferences, eventually culminating in a deal.

That is, each business negotiation includes one or more web conferences. Also, each business negotiation may include multiple web conferences that are held consecutively in the same timeline. In this case, one or more web conferences held after a certain web conference are referred to as subsequent web conferences.

Figure 1 shows an overview of an embodiment. In the following explanation, "UI" is an abbreviation for User Interface, which typically refers to a GUI (Graphical User Interface).

The web conference support system includes an interface device (not shown), a storage device 10, and a processor (not shown) connected to them.

The storage device 10 stores at least a web conference support program. The web conference support program is a computer program for supporting web conferences.

When the web conference support program is executed by the processor, processes such as video analysis, conference information extraction, reference information extraction, and recommendation target extraction are performed. The details of these processes are described below, but referring to Figure 1, the outline is as follows.

During business negotiations, a web conference between a sales representative and a customer is recorded, and the recorded data is stored in the data storage unit 104. The web conference support system performs video analysis processing, which is an analysis process of the recorded data. The web conference support system performs conference information extraction processing, which extracts conference information based on the results of the video analysis.

The extracted conference information is stored in the data storage unit 104. Each web conference that constitutes a business negotiation includes one or more topics. Furthermore, a topic is composed of a collection of one or more utterances whose contents are highly related to each other. That is, each topic includes one or more utterances by at least one customer and one sales representative. The web conference support system calculates an attention level, which is an evaluation value based on at least the facial expressions, actions, and utterances of the customer recorded in the recorded data, and includes the attention level in the conference information to which the recorded data is associated. The data storage unit 104 stores recorded data of multiple web conferences that were held consecutively during the same business negotiation (an example of the same time series) for at least one business negotiation.

When the web conference support system receives a piece of recorded data from a sales representative, it performs a reference information extraction process. That is, the web conference support system acquires recorded data from the data storage unit 104 that has the same topic type as the designated recorded data and has a difference in attention level with the designated recorded data within a predetermined range, and displays reference information, which is information about the acquired recorded data, for the sales representative. The reference information may include information about the recorded data of at least one subsequent web conference held after the web conference represented by the acquired recorded data.

Furthermore, when the web conference support system identifies customer information based on the acquired recording data, the web conference support system may perform a recommendation target extraction process to identify a recommendation target (e.g., a product or service) based on the customer information. The web conference support system may display information related to the identified recommendation target to a sales representative. This embodiment will be described in detail below.

Figure 2 shows an example of the overall system configuration including the web conference support system 100.

Note that each block described below represents a functional block, not a hardware configuration.

The web conference support system 100 includes a data processing unit 102, a data storage unit 104, a user interface unit 106, and a communication unit 108.

The user interface unit 106 accepts operations from the user via the touch panel, and is responsible for processing related to the user interface, such as image display and audio output. The communication unit 108 is responsible for communication with other devices, such as servers and terminals, via the Internet 290 (an example of a communication network). The data storage unit 104 stores various data. The data processing unit 102 executes various processes based on data acquired by the user interface unit 106 and the communication unit 108, and data stored in the data storage unit 104. The data processing unit 102 also functions as an interface for the user interface unit 106, the communication unit 108, and the data storage unit 104.

The data processing unit 102 includes a video analysis unit 121, a meeting information extraction unit 122, a reference information extraction unit 123, and a recommendation target extraction unit 124.

The video analysis unit 121 analyzes the recorded data of the web conference, which is a video. The process in which the video analysis unit 121 analyzes the recorded data of the web conference is referred to as video analysis processing. Details of the video analysis processing will be described later in relation to FIG. 10.

Based on the results of the video analysis process performed by the video analysis unit 121, the conference information extraction unit 122 generates conference information data, which is data indicating various attribute information related to the web conference represented by the recorded data to be analyzed. The process in which the conference information extraction unit 122 generates conference information data is referred to as a conference information extraction process. Details of the conference information extraction process will be described later in relation to FIG. 11.

The meeting information extraction unit 122 includes a customer type classification unit 201, a topic segmentation unit 202, an emotion ratio calculation unit 203, an attention calculation unit 204, an important/NG word calculation unit 205, and a usage material extraction unit 206.

The customer type classification unit 201 classifies the target customer into a corresponding customer type among multiple customer types. The topic division unit 202 identifies the topic raised in the web conference represented by the target recorded data based on the result of the video analysis process performed by the video analysis unit 121 (for example, divides the recorded data by topic). The emotion ratio calculation unit 203 calculates the emotion ratio of the customer during the conference based on the result of the video analysis process performed by the video analysis unit 121. The attention calculation unit 204 calculates the attention level of the target recorded data based on the result of the video analysis process performed by the video analysis unit 121. The attention level is an evaluation value that quantitatively evaluates the interest level of at least one participant (for example, a customer) based on the facial expression, action, and speech content of the participant recorded in the target recorded data. The important/NG word calculation unit 205 calculates important words and/or NG words based on the result of the video analysis process performed by the video analysis unit 121. The important word is a word that has increased the attention level as a result of being used in the web conference represented by the target recorded data. Moreover, the NG words are words that, when used in the web conference represented by the target recorded data, result in a decrease in the level of attention. The used material extraction unit 206 extracts the used materials based on the results of the video analysis process performed by the video analysis unit 121.

The reference information extraction unit 123 extracts reference information based on the results of the conference information extraction process performed by the conference information extraction unit 122 and each of the functional units (201 to 206) contained therein. The reference information is information presented to the user to assist in achieving the purpose of the web conference, and specifically includes at least one of the video represented by the recorded data and the conference information represented by the conference information data. Details of the reference information extraction process will be described later in relation to FIG. 13.

The reference information extraction unit 123 also performs a difference point calculation process. The difference point is an evaluation value for determining the order in which to present reference information to the user when multiple pieces of reference information exist. Details of the reference information extraction process will be described later with reference to FIG. 14.

The recommendation target extraction unit 124 extracts recommendation targets. A recommendation target is a product or service that is likely to be appealing when proposed to a customer who is a participant of a web conference. Details of the recommendation target extraction process will be described later in relation to FIG. 19.

The user interface unit 106 includes an input unit 161 that accepts input from the user, and an output unit 162 that outputs various information such as videos, images, documents, etc. to the user.

The correspondence between the components shown in FIG. 2 and the components shown in FIG. 1 is, for example, as follows. That is, the video analysis process shown in FIG. 1 is executed by the video analysis unit 121 shown in FIG. 2. The meeting information extraction process shown in FIG. 1 is executed by each of the functional units of the customer type classification unit 201, topic division unit 202, emotion ratio calculation unit 203, attention calculation unit 204, important/NG word calculation unit 205, and use material extraction unit 206 included in the meeting information extraction unit 122 shown in FIG. 2. The reference information extraction process shown in FIG. 1 is executed by the reference information extraction unit 123 shown in FIG. 2. The recommendation target extraction unit 124 shown in FIG. 2 executes the recommendation target extraction process shown in FIG. 1.

The data storage unit 104 includes a recording data storage unit 141, a video analysis result storage unit 142, an emotion type data storage unit 143, a meeting information data storage unit 144, and a customer information data storage unit 145.

The recording data storage unit 141 stores the recording data for each web conference. The video analysis result storage unit 142 stores video analysis result data, which is data representing the results of the video analysis process performed by the video analysis unit 121. The emotion type data storage unit 143 stores emotion type data, which is a definition collection that defines the correspondence between the facial expressions of speakers during web conferences and specific emotion types. The conference information data storage unit 144 stores the conference information data generated by the conference information extraction unit 122. The customer information data storage unit 145 stores customer information, which is information regarding the attributes of customers.

The functional units included in the data storage unit 104, namely the recording data storage unit 141, the video analysis result storage unit 142, the emotion type data storage unit 143, the meeting information data storage unit 144, and the customer information data storage unit 145, each store data represented by the tables illustrated in Figures 3 to 7.

The tables shown in Figures 3 to 7 are explained below.

Figure 3 shows an example of the configuration of a video management table stored in the recording data storage unit 141.

The video management table 300 is a table for managing the recorded data for each web conference. The video management table 300 has a record for each recorded data. The record represents a video ID, which is an identification number of the recorded data, the start date and time and the end date and time of the web conference represented by the recorded data, the names of the customers and sales representatives who participated in the web conference (customer name and representative name), a business negotiation ID, which is an identification number for managing the business negotiation that is the purpose of the web conference, the number of business negotiations that indicates how many times the web conference is a meeting out of the total number of business negotiations, and the topics that were discussed in the web conference. In other words, the video management table 300 roughly represents the contents of the web conference represented by each recorded data by linking and recording a video ID that identifies the recorded data with various components of the web conference represented by the recorded data for each record. In the example shown in FIG. 3, the video management table 300 records, as components of the web conference represented by the recorded data assigned the video ID "220106001," that it is the first meeting in the entire business negotiation assigned the business negotiation ID "0001," that it started at 14:00 on January 6, 2022 and ended at 15:40, that customer A and sales representative P were participants, and that four topics, A, B, C, and D, were discussed at the meeting.

The method for extracting topics from each recorded data will be described later in relation to Figure 11.

Figure 4 shows an example of the configuration of a video analysis result table stored in the video analysis result storage unit 142.

The video analysis result table 400 is a table for recording video analysis result data, which is data showing the results of video analysis processing by the video analysis unit 121, for each recorded data. For ease of explanation, Fig. 4 shows a schematic table showing the results of video analysis processing for any one recorded data. That is, the video analysis result table 400 shown in Fig. 4 exists for each recorded data. The video analysis result table 400 has a record for each utterance during a web conference represented by each recorded data. The record includes an utterance ID, which is an identification number for identifying each utterance, a time indicating the time period in the recorded data to be analyzed in which each utterance was made, the utterance content recorded by converting the specific content of each utterance into text format, individual words extracted from the utterance content, positive words having positive attributes among the extracted words, negative words having negative attributes among the extracted words, the speaker of each utterance, the emotion of the speaker when each utterance was made, and the behavior such as gestures and gestures shown by the speaker when each utterance was made. That is, the video analysis result table 400 represents the content of each utterance during a web conference represented by each recorded data in detail by linking and recording the utterance ID that identifies the utterance and various components of the utterance for each record. According to the example shown in FIG. 4, the video analysis result table 400 records data indicating the following items for an utterance assigned an utterance ID of "220106001-005".
- The utterance in question was made by a customer who was a participant in the web conference during the time period from 2 minutes 40 seconds to 2 minutes 55 seconds from the start of the recorded data being analyzed, and was, "I see. When I hear things like that, it's important to be prepared after all."
- The utterance in question contains four words: "I see,""To hear that,""Beprepared," and "It's important," and of these, the word "I see" has positive attributes.
- The customer made this utterance while nodding and expressing an emotion classified as "Surprise."

Figure 5 shows an example of the configuration of an emotion type table stored in the emotion type data storage unit 143.

The emotion type table 500 is a definition collection that defines in advance the correspondence between various facial expressions and various emotion types, and is a table for recording emotion type data that the video analysis unit 121 refers to when performing video analysis processing. The emotion type table 500 has a record for each facial expression. The record represents an image that represents a typical type of various facial expressions, each type of facial expression, each type of emotion, and the attribute to which each emotion type is classified. That is, the emotion type table 500 records, for each record, an image that represents a typical type of various facial expressions, the type of the facial expression, the type of emotion that corresponds to the facial expression type, and the attribute of the emotion type in association with each other. According to the example shown in FIG. 5, the video management table 300 defines that when the facial expression of a speaker during speech is included in the facial expression type of "smile", the emotion of the speaker during speech is "happy", and the emotion type of "happy" is classified into a positive emotion attribute.

Figure 6 shows an example of the configuration of a conference information management table stored in the conference information data storage unit 144.

The conference information management table 600 is a table for recording the conference information data extracted by the conference information extraction unit 122 through the conference information extraction process. For ease of explanation, FIG. 6 shows a schematic diagram of a table for a web conference represented by any one piece of recorded data. That is, the conference information management table 600 illustrated in FIG. 6 exists for each web conference represented by the recorded data, that is, for each piece of recorded data. The records of the conference information management table 600 represent the video ID of the corresponding recorded data, the name, various attributes and type of the customer who participated in the web conference, the proposal content to the customer, the current status of the business negotiation that is the purpose of the web conference, frequent words that represent words that frequently appeared during the web conference, important words and NG words during the web conference, topics that were discussed in the web conference, and materials used that are materials used for explanations in the web conference. The video ID is also used as an identification number for identifying the web conference represented by the recorded data.

In the fields of each record where data related to customer attributes is recorded, data representing various customer attributes such as gender and age is recorded for each attribute. This data representing various customer attributes is the same as the data recorded in the customer management table 700 described below.

In each record, the field in which data related to the proposal is recorded contains the name and type of the specific product or service.

In each record, the field in which data on topics is recorded contains, for each topic, the time period when the topic was discussed, the percentage of the customer's emotions regarding that topic during the meeting, and the customer's level of attention regarding that topic during the meeting. The percentage of emotions refers to the percentage of the total time that the customer felt each emotion during the web conference as a percentage of the total time the web conference was held.

In the field of each record where data on the materials used is recorded, the number of slides that make up the material, the length of time the material was presented to the customer, and the level of attention the customer received while the material was being presented are recorded for each material.

That is, the conference information management table 600 centrally records, for each record, various information related to the web conference extracted as a result of the video analysis process by the video analysis unit 121 and the conference information extraction process by the conference information extraction unit 122, in association with a video ID that identifies the web conference, thereby representing the details of each web conference. According to the example shown in Fig. 6, the conference information management table 600 centrally records data representing the following items for a web conference identified by a video ID of "220106001".
- The web conference in question constituted a business negotiation to propose level term insurance, a type of term insurance, to a young female customer named A.
- The business negotiations that were the purpose of the web conference were ultimately concluded through that web conference or a subsequent web conference.
- During the web conference, words such as "money,""insurance,""medicalexpenses,""benefits,""cancer,""risk,""coverage,""health," and "cancellation" were frequently used.
-Key words used in the web conference included "money,""medicalcosts,""benefits," and "cancer."
- The prohibited words used in the web conference were "risk,""health,""cancellation," etc.
- During the web conference, at least two topics, A and B, were discussed.
- The attention rate for the period from the start of the web conference up to 3 minutes and 20 seconds when Topic A was on the table for the meeting was 30%, and the proportion of customers' emotions during that period was blank 50%, smiling 31%, surprised 2%, satisfied 17% and anger 0%.
- During the web conference, at least three types of materials, AAAA, BBBB, and CCCC, were used for explanations.
- Of these, for document AAAA, 12 pages of slides were presented to the client for 21 minutes and 15 seconds, during which time the attention rate was 46%.

In this embodiment, the conference information management table 600 is described as having a column for recording frequently used words during a web conference, but the conference information management table may be configured without including this column. Also, each record in the conference information management table may include one or more fields for recording other elements that may constitute the conference information.

Figure 7 shows an example of the configuration of a customer management table stored in the customer information data storage unit 145.

The customer management table 700 is a table for managing customer information data of customers who are the subject of each business negotiation. The customer management table 700 has a record for each customer. The record indicates the customer's name, sex, age, whether they have a spouse, the number of children, their employment status, their annual income for the previous year, and their customer type. In other words, the customer management table 700 records, for each record, the customer name that identifies the customer and data that indicates the various attributes of the customer, thereby roughly representing the personality of each customer. According to the example shown in FIG. 7, the customer management table 700 records that a customer named A is a 23-year-old woman with a spouse and one child, is an office worker with an annual income of 5 million yen for the previous year, and is an "aimable" customer type.

Figure 8 shows an example of a customer type configuration.

Customer types classify customers' personalities into four predefined types by combining two evaluation indices: "emotional expressiveness," which indicates the richness of emotional expression, and "self-assertiveness," which indicates the level of self-assertiveness.

According to the example shown in Figure 8, customers with low emotional expression and high self-assertiveness are classified as "Driving", customers with high emotional expression and high self-assertiveness are classified as "Expressive", customers with high emotional expression and low self-assertiveness are classified as "Amiable", and customers with low emotional expression and low self-assertiveness are classified as "Analytical".

As described above, data representing the customer type of each customer is recorded as elements in the conference information management table 600 and the customer information table 700. Of these, the contents of the conference information management table 600 stored in the conference information data storage unit 144 are presented to the user as visual information for each record, as described below. Depending on the difference in the customer type of the corresponding customer, even if the same utterance is made in the same situation or the same materials are presented, the reaction may be significantly different. However, with the web conference support system 100, the user can appropriately proceed with the business negotiations while keeping in mind the customer type of the customer in charge. As a result, the user can easily achieve the purpose of the web conference.

The method for classifying customer types will be described later in relation to Figure 11.

The above is an explanation of examples of tables stored in each of the functional units, the recording data storage unit 141, the video analysis result storage unit 142, the emotion type data storage unit 143, the meeting information data storage unit 144, and the customer information data storage unit 145.

In this embodiment, the web conference support system 100 has been described as including the emotion type data storage unit 143 and the customer information data storage unit 145. However, the web conference support system may be configured without including the emotion type data storage unit 143 and the customer information data storage unit 145. In this case, the web conference support system may obtain various data represented by the customer management table 700 and the emotion type table 500 from another device (not shown) via the communication unit 108 and execute each process.

In addition, in this embodiment, the conference information management table 600 has been described as including overlapping customer information recorded in the customer management table 700. However, the customer information does not need to be included in the recording targets of the conference information management table 600. In this case, when presenting the conference information to the user as visual information, the web conference support system may generate conference information data by having the conference information extraction unit 122 acquire data from the conference information data storage unit 144 and the customer information data storage unit 145.

In this embodiment, the functions of the web conference support system 100 have been described as being realized integrally by a single computer device, but these functions may also be realized by multiple computer devices or server devices connected to each other. Also, the configuration may include a general-purpose computer such as a laptop PC with a web browser installed thereon, or may include a web server.

An example of the processing performed in this embodiment is described below.

Figure 9 shows an example of the overall processing flow performed in this embodiment.

In this embodiment, when the web conference support system 100 receives a video specification from a user (e.g., a sales representative) via the input unit 161 or the communication unit 108 (S901), the web conference support system 100 retrieves the recorded data of the specified video from the recorded data storage unit 141 and analyzes it (S902). The video analysis process described above refers to the process performed in S902. Details of the video analysis process will be described later with reference to FIG. 10.

Next, the web conference support system 100 extracts conference information based on the video analysis result data acquired from the video analysis result storage unit 142 (S903). The conference information extraction process described above refers to the process performed in S903. Details of the conference information extraction process will be described later in relation to FIG. 11.

The conference information extraction process performed in S903 includes an attention calculation process. Details of the attention calculation process will be described later with reference to FIG. 12.

Next, the web conference support system 100 extracts reference information based on the conference information acquired from the conference information data storage unit 144 (S904), and presents the extraction result to the user as visual information via the output unit 162 or the communication unit 108 (S905). The above-mentioned reference information extraction process refers to the process performed in S904. Details of the reference information extraction process will be described later in relation to FIG. 13.

Other examples of processes performed by the web conference support system 100 in this embodiment include a difference point calculation process and a recommendation target extraction process. The difference point calculation process is a process for calculating difference points used to determine the ranking of multiple pieces of reference information. The recommendation target extraction process is a process for extracting a recommendation target to be proposed to a web conference customer. Details of the difference point calculation process will be described later in relation to FIG. 14. Details of the recommendation target extraction process will be described later in relation to FIG. 19.

Figure 10 shows an example of the flow of video analysis processing.

The video analysis unit 121 acquires the recording data to be analyzed from the recording data storage unit 141 (S1001) and extracts the customer's emotion information recorded in the recording data (S1002).

Emotion information is information that indicates the emotions felt by the customer during the web conference. When the speaker is a customer, it indicates the emotions felt by the customer while he or she was speaking. When the speaker is a sales representative, it indicates the emotions felt by the customer while the sales representative was speaking. In other words, the extracted customer emotion information is linked to information about each utterance and stored in the video analysis result storage unit 142, regardless of whether the speaker is a customer or a sales representative.

Emotion information can be extracted, for example, by extracting the facial expressions and movements of a customer during a web conference using a known video analysis method, and then appropriately classifying the extraction results by referring to the emotion type data stored in the emotion type data storage unit 143. At this time, the video analysis unit 121 may quantitatively evaluate the extracted facial expressions and movements of the customer according to predetermined criteria. The video analysis unit 121 may also extract emotional information of the customer from the voice uttered by the customer. In this case, the video analysis unit 121 may analyze the audio information included in the recorded data using a known audio analysis method.

Next, the video analysis unit 121 converts the speaker's speech content into text format for each utterance using a known voice analysis method (S1003). At this time, the video analysis unit 121 identifies the speaker for each utterance and assigns speaker information for that utterance to each piece of text data representing the speech content.

Next, the video analysis unit 121 extracts the words used by the speaker during each utterance using methods such as morphological analysis (S1004).

Then, the video analysis unit 121 generates video analysis result data based on the results of each process from S1002 to S1004 (S1005), stores it in the video analysis result storage unit 142, and completes the video analysis process.

As a result of the video analysis process, the emotional information of the customer and the content of the speaker's speech recorded in the recorded data are extracted in a data format that is easy to analyze using known data analysis methods. This allows the web conference support system 100 to execute the conference information extraction process described below.

In addition, as a result of the video analysis process, the speaker is identified for each utterance. This allows only the customer's utterances to be extracted from all utterances made during the web conference, allowing the web conference support system 100 to perform the attention calculation process described below.

Figure 11 shows an example of the flow of the meeting information extraction process.

The meeting information extraction unit 122 acquires the video analysis result data to be processed from the video analysis result storage unit 142 (S1101). At this time, the meeting information extraction unit 122 also refers to the recording data storage unit 141 and acquires data representing the customer name from the corresponding record in the video management table 300.

Next, the customer type classification unit 201 classifies the customer into one of four customer types based on the video analysis results (S1102).

The customer type classification method is as follows. First, the customer type classification unit 201 calculates the emotional expression level of the customer based on the customer's emotional information included in the video analysis result data. As a result, if the emotional expression rate, which is the percentage of the total time during which the customer clearly expressed his/her emotions during the web conference, is less than 30%, the customer is determined to be a type that suppresses emotional expression. On the other hand, if the emotional expression rate is 30% or more, the customer is determined to be a type that expresses a lot of emotions. Next, the customer type classification unit 201 calculates the customer's assertiveness based on the speaker's speech information included in the video analysis result data. As a result, if the speech rate, which is the percentage of the speech time of the customer's speech in the total speech time of all speeches made in the web conference, is less than 30%, the customer is determined to be a type that has a low level of assertiveness and listens to the opinions of others. On the other hand, if the speech rate is 30% or more, the customer is determined to be a type that strongly asserts himself/herself.

In this embodiment, the thresholds for determining the emotional expressiveness and the self-assertiveness are each described as 30%, but the criteria for determining the emotional expressiveness and the self-assertiveness may be changed as appropriate.

Next, the topic segmentation unit 202 classifies the speech content into clusters for each topic based on the speaker's speech information included in the video analysis result data (S1103). Each topic can be extracted, for example, by analyzing the occurrence rate of words in the speech and the degree of co-occurrence between words, and comparing the analysis results with a well-known thesaurus database. In addition, the speech content can be classified into clusters for each topic by, for example, comparing the occurrence rate of words in the speech and the degree of co-occurrence between words for each utterance, and dividing the speech content into multiple clusters before and after a point where the topic changes.

Next, the emotion ratio calculation unit 203 calculates the emotion ratio of the customer based on the emotion information of the customer contained in the video analysis result data (S1104). The emotion ratio can be calculated, for example, by calculating the percentage of the total time that each emotion is expressed for each emotion shown by the customer during the web conference, and then referring to the emotion type table 500, assuming that the calculation result for each emotion is the calculation result for the emotion that corresponds to the type of the emotion.

Next, the attention level calculation unit 204 calculates the attention level based on the customer's emotion information and the speaker's speech information contained in the video analysis result data (S1105). The attention level calculation process described above refers to the process performed in S1105 (described in detail later in relation to FIG. 12).

Next, the important/NG word extraction unit 205 extracts at least one of important words and NG words in the web conference based on the customer's emotional information and the speaker's speech information contained in the video analysis result data (S1106). Important words can be extracted, for example, by calculating the attention level for each topic or utterance, identifying the word used immediately before the timing when the attention level increased, and verifying the reproducibility of the identified word's contribution to the increase in attention level. Similarly, NG words can be extracted, for example, by calculating the attention level for each topic or utterance, identifying the word used immediately before the timing when the attention level decreased, and verifying the reproducibility of the identified word's contribution to the decrease in attention level.

In relation to S1106, the important/bad word extraction unit 205 may analyze the speaker's speech information contained in the video analysis result data and extract frequent words based on the appearance rate of words in the speech content. In this case, the important words and bad words may be extracted by a method of narrowing down the extracted frequent words on the assumption that the frequent words, important words, and bad words are in an inclusive relationship with each other, or the frequent words, important words, and bad words may be extracted separately from each other on the assumption that they are unrelated to each other.

Next, the used material extraction unit 206 extracts the used materials used during the web conference (S1107). The used materials may be extracted by analyzing the recorded data using well-known techniques such as video analysis and audio analysis, or the used material extraction unit 206 may refer to the system log of the terminal used in the web conference via the communication unit 108. The used material extraction unit 206 may also be capable of accepting a registration operation of the used materials from a user or a system administrator via the input unit 161 or the communication unit 108.

Then, the conference information extraction unit 122 generates conference information data based on the results of each process from S1101 to S1107, stores the data in the conference information data storage unit 144, and completes the conference information extraction process.

Figure 12 shows an example of the flow of the attention calculation process.

The attention calculation unit 204 first calculates the proportion of facial expressions classified as positive attributes and the proportion of facial expressions classified as negative attributes among all facial expressions shown by the customer during the web conference, based on the duration of each facial expression (S1201), and converts the calculation results into points as follows (S1202). For example, if the total time during which facial expressions classified as positive attributes were shown accounted for 20% of the total conference time, the points related to facial expressions are evaluated as "20 points." Similarly, if the total time during which facial expressions classified as negative attributes were shown accounted for 5% of the total conference time, the points related to facial expressions are evaluated as "-5 points."

Next, the attention calculation unit 204 calculates the number of actions classified as positive attributes and the number of actions classified as negative attributes performed by the customer during the web conference (S1203), and converts the calculation results into points as follows (S1204). For example, if the customer performs five actions classified as positive attributes during the web conference, the points for the actions are "25", which is the product of the constant "+5", which represents the points per action, multiplied by the variable "5", which represents the number of actions. Similarly, if the customer performs two actions classified as negative attributes during the web conference, the points for the actions are "-10", which is the product of the constant "-5", which represents the points per action, multiplied by the variable "2", which represents the number of actions.

Next, the attention calculation unit 204 calculates the number of times the customer used words classified as positive attributes and words classified as negative attributes during the web conference (S1205), and converts the calculation results into points in the following manner (S1206). For example, if the customer used a word classified as positive attributes four times during the web conference, the points for the word are "20", which is the product of the constant "+5", which represents the points per use, multiplied by the variable "4", which represents the number of uses. Similarly, if the customer used a word classified as negative attributes once during the web conference, the points for the word are "-5", which is the product of the constant "-5", which represents the points per use, multiplied by the variable "1", which represents the number of uses.

After completing each process of S1201 to S1206, the attention calculation unit 204 calculates the attention level by adding up all the points related to facial expressions, actions, and words (S1207). For example, if the points related to facial expressions are "20" and "-5", the points related to actions are "25" and "-10", and the points related to words are "20" and "-5", the attention level is "45". After S1207 ends, the attention calculation unit 204 completes the attention level calculation process.

In this embodiment, the constants representing the points per action count and per use of a word are "+5" or "-5", but these constants may be changed as appropriate. The method of calculating each point may also be changed as appropriate.

In the present embodiment, the attention level is calculated for each recorded data, i.e., for each web conference represented by the recorded data. However, the attention level may be calculated for each topic during the web conference, or for each utterance during the web conference. The attention level may also be calculated for each word or phrase included in an utterance. For example, FIG. 6 used to explain the conference information management table 600 shows an example of the attention level calculated for each topic, with the attention level being "30%" during the time when topic A was on the table, and the attention level being "52%" during the time when topic B was on the table. For example, the explanation of the method of extracting important words and NG words described above in relation to FIG. 11 is an example of the attention level being calculated for each topic or for each utterance.

When calculating the attention level for a larger object in an inclusion relationship, such as a topic for an utterance and a web conference, or a web conference for a topic, for example, each process from S1201 to S1207 may be performed for each of the relatively small units of calculation objects (e.g., utterances), and the average or sum of the calculated attention levels may then be used as the attention level for the relatively large calculation object (e.g., a topic).

Furthermore, in this embodiment, as illustrated in FIG. 11, the attention calculation process is described as being included in the conference information extraction process, but the attention calculation process does not have to be included in the conference information extraction process. In this case, the web conference support system may execute only the attention calculation process without executing the conference information extraction process.

As a result of the attention calculation process, the customer's attention level in the web conference is calculated. The attention level is an evaluation value that quantitatively evaluates the customer's level of interest based on the customer's facial expressions, actions, and speech during the web conference. The web conference support system 100 extracts recording data and conference information data of other web conferences as reference information based on at least two quantitative evaluation criteria: the degree of similarity of attention levels and the degree of similarity of topics. In other words, this allows the web conference support system 100 to execute a process for extracting reference information (described in detail below in relation to FIG. 13).

If reference information is extracted based only on the degree of similarity of topics, for example, even if the topics discussed during the meeting are the same, such as "worry," "preparation," "insurance," and "contracts," there may be web conferences where progress is being made toward a deal and web conferences where negotiations have broken down, and so on, the information extracted as reference information may be a mixture of web conference contents and trends. However, if reference information is extracted based on the degree of attention in addition to the degree of similarity of topics, it becomes possible to extract reference information that is more in line with the user's needs. In other words, this makes it possible to accurately extract reference information based on two quantitative evaluation criteria, making it easier for the web conference support system 100 to provide web conference support that contributes to achieving the purpose of the web conference.

In addition, as a result of the conference information extraction process, conference information data of the web conference represented by each recorded data is extracted. The conference information data represents the content of the corresponding web conference in detail, and is presented to the user as visual information at the user's request. This allows the user to grasp in detail the content of the web conference represented by the presented conference information data, making it easier to achieve the purpose of the web conference by using the presented information as future reference.

Figure 13 shows an example of the flow of the reference information extraction process.

When the reference information extraction unit 123 receives a designation of a video or meeting information from the user via the input unit 161 or the communication unit 108 (S1301), it determines whether the designated target is meeting information (S1302). If the designated target is meeting information (S1302: YES), it obtains meeting information data representing the designated meeting information from the meeting information data storage unit 144 (S1304).

On the other hand, if the specified target is not conference information (S1302: NO), the specified target is a video, i.e., recorded data. In this case, conference information related to the web conference represented by the specified recorded data is automatically specified (S1303), and the conference information data represented by the automatically specified conference information is obtained from the conference information data storage unit 144 (S1304).

Next, the reference information extraction unit 123 extracts one or more pieces of conference information data whose contents are similar to the acquired conference information data from the conference information data storage unit 144, based on at least the degree of topic similarity and the degree of attention approximation (S1305).

Next, the reference information extraction unit 123 acquires a video corresponding to the extracted conference information data, i.e., recorded data, from the recorded data storage unit 141 (S1306). After S1306 is completed, the reference information extraction unit 123 presents at least one of the acquired conference information data and recorded data as reference information to the user via the output unit 162 or the communication unit 108 (S1307). After S1307 is completed, the reference information extraction unit 123 completes the reference information extraction process.

Note that, for the recording data and conference information data extracted as a result of each process of S1301 to S1306, if a subsequent web conference that constitutes the same business negotiation as the web conference represented by these data has already been held, the reference information extraction unit 123 may obtain at least one of the recording data and conference information data representing the subsequent web conference that has already been held from the recording data storage unit 141 or the conference information data storage unit 144, and present it to the user as reference information.

If there are multiple subsequent web conferences that have already been held, recording data, etc. may be acquired for all of them, or recording data, etc. may be acquired only for the web conference that was held the earliest among the subsequent web conferences. In such a case, the user may be able to select which of the multiple subsequent web conferences the recording data, etc. of which is to be acquired. In this case, the number of selectable web conferences may be set arbitrarily.

In the processes of S1305 to S1307, the reference information extraction unit 123 extracts the recording data and the meeting information data to be presented to the user as reference information based on at least the degree of similarity of the topics and the degree of attention.

For example, a threshold is set for the number of times the same topic is included as in the specified recorded data or meeting information data, and a judgment process is performed to determine the degree of topic similarity. To explain this using the contents of the video management table 600 illustrated in FIG. 6 as an example, if a user specifies recorded data with video ID "220106001" and the number of topics set as the criterion for judging the degree of topic similarity is "3", the recorded data extracted as reference information by the reference information extraction unit 123 will be the two video IDs "220222013" and "220106002".

In addition, for example, the degree of approximation of the attention level is determined by setting an allowable range for the difference in attention level with respect to the attention level of the recording data or conference information data specified by the user, and performing a process of determining whether the difference in attention level is within the allowable range. If the user specifies conference information data with an attention level of "45", the thresholds for determining that the difference in attention level is within the allowable range are "+5" and "-5", and there are five other conference information data with attention levels of "48", "37", "64", "12", and "44", respectively, the two conference information data that the reference information extraction unit 123 extracts as reference information are the conference information data with an attention level of "48" with an attention level difference of "+3", and the conference information data with an attention level of "44" with an attention level difference of "-1".

In this embodiment, it has been described that whether the contents of other arbitrary recorded data and meeting information data are similar to the recorded data or meeting information data specified by the user is determined based on at least the degree of topic agreement and the degree of attention. However, at least one of the recorded items of each record of the video management table 300 illustrated in FIG. 3 or the meeting information management table 600 illustrated in FIG. 6 may be added to the criteria for determining the similarity between other arbitrary recorded data and meeting information data and the recorded data or meeting information data specified by the user. In this case, the reference information extraction unit 123 may refer to the recorded data storage unit 141 or the meeting information data storage unit 144 to appropriately acquire data and use it to determine the similarity.

In addition, when presenting multiple recorded data or meeting information data to a user as reference information, the reference information extraction unit 123 may calculate an evaluation value of the difference in attention level between two pieces of data, called a difference point, for each recorded data, etc. to be presented, and present multiple recorded data, etc. to the user in order of the smallest difference point.

Figure 14 shows an example of the flow of the difference point calculation process.

Note that Figure 14 illustrates an example of processing when both sets of data to be compared are conference information data, and the level of attention is calculated for each topic.

The reference information extraction unit 123 first selects one piece of conference information data for another conference (S1401).

Next, the reference information extraction unit 123 extracts one topic from the conference information data specified by the user (S1402), extracts one topic from the conference information data of another conference (S1403), and calculates the word matching rate between the two topics (S1404).

Next, if there are unprocessed topics remaining in the conference information data of other conferences (S1405: YES), the reference information extraction unit 123 performs the processes of S1403 to S1404 again. This process is repeated until there are no unprocessed topics remaining in the conference information data of other conferences.

If there are no unprocessed topics remaining in the conference information data of the other conference (S1405: NO), the difference in attention between the two topics with the highest word match rate is converted into points (S1406). For example, if the conference information data of the other conference includes three topics, topic A, topic B, and topic C, and the word match rate with topic P in the conference information data specified by the user is 80% for topic A, 45% for topic B, and 15% for topic C, the reference information extraction unit 123 selects topic A. Then, if the attention level for topic A in the conference information data of the other conference is "40" and the attention level for topic P in the conference information data specified by the user is "30", the reference information extraction unit 123 calculates the difference between the two topics as "10".

Next, if there are unprocessed topics remaining in the conference information data specified by the user (S1407: YES), the reference information extraction unit 123 performs the processes of S1402 to S1406 again. This process is repeated until there are no unprocessed topics remaining in the conference information data. For example, if the conference information data specified by the user includes three topics, topic P, topic Q, and topic R, the reference information extraction unit 123 performs the processes of S1402 to S1406 for all of topic P, topic Q, and topic R.

If there are no unprocessed topics remaining in the conference information data (S1407: NO), the differential points are calculated (S1408). The differential points are calculated by dividing the sum of the attention level differences for each topic by the total number of topics included in the conference information data. For example, if the attention level difference between topic P and topic A is "10", the attention level difference between topic Q and topic B is "30", and the attention level difference between topic R and topic C is "20", the differential points are "(10 + 30 + 20) ÷ 3 = 20 points".

If there is more than one conference information data for other conferences, the reference information extraction unit 123 performs the processes S1401 to S1408 for each of them. The processes S1401 to S1408 are repeated until there is no unprocessed conference information data for other conferences.

After completing the processes of S1401 to S1408 for the conference information data of all other conferences to be processed, the reference information extraction unit 123 sorts the conference information data in ascending order of differential points and presents it to the user as reference information (S1409). For example, if there are two conference information data, conference information data α with a differential point of "20 points" and conference information data β with a differential point of "15 points", the reference information extraction unit 123 sorts them in the order of conference information data β and conference information data α, and presents them to the user via the output unit 161 or the communication unit 108.

After S1409 is completed, the reference information extraction unit 123 completes the difference point calculation process.

As a result of the difference point calculation process, a difference point between each of the multiple recorded data or conference information data presented to the user as reference information and the recorded data or conference information data specified by the user is calculated. The difference point is a value that represents the difference in attention level between two pieces of data, and the lower the difference point, the more similar the recorded data is to the recorded data specified by the user. In other words, when there are multiple recorded data to present to the user as reference information, the web conference support system 100 can present to the user the recorded data with higher similarity on a priority basis.

Figure 15 shows an example of a reference information display screen.

As a result of the reference information extraction process, various information related to the web conference represented by the recorded data is presented to the user, as shown in FIG. 15. In the case of FIG. 15, three pieces of conference information data identified by video IDs "220218001", "150726001", and "200726001" are displayed as reference information in ascending order of difference points. When the user performs a selection operation on each of the character strings representing these conference information data, the screen transitions and the detailed contents of the selected conference information data are displayed, for example, as shown in FIG. 6. In addition, if the web conference represented by these conference information data has one or more subsequent web conferences, the reference information extraction unit 123 can also display the detailed contents of the conference information data for the subsequent web conference as reference information.

In addition, since the conference information data and the recording data of the web conference that the conference information data represents are linked to each other and stored in the recording data storage unit 141 and the conference information data storage unit 143, the reference information extraction unit 123 can also display the corresponding recording data as reference information.

As described above, the recorded data or conference information data extracted as reference information is extracted with high accuracy based on at least two quantitative evaluation criteria, the degree of topic similarity and the degree of attention, and therefore there is a high possibility that the contents are very similar to the recorded data or conference information data specified by the user. Therefore, when there are one or more subsequent web conferences to the web conference represented by the extracted recorded data or conference information data, information on the process and outcome of the business negotiations by those multiple web conferences is useful when making advance preparations for the subsequent web conference that has not yet been held after the web conference represented by the recorded data or conference information data specified by the user. Therefore, when making advance preparations for the subsequent web conference that has not yet been held after the web conference represented by the specified recorded data or conference information data, the user can easily achieve the purpose of the web conference by appropriately referring to the contents of the recorded data or conference information data extracted by the reference information extraction unit 123. As a result, the web conference support system 100 can provide web conference support that contributes to achieving the purpose of the web conference.

The above embodiments can be summarized as follows:

The web conference support system 100 includes a recording data storage unit 141, an attention level calculation unit 204, a reference information extraction unit 123, a conference information data storage unit 143, an input unit 161, an output unit 162, and a communication unit 108. The recording data storage unit 141 stores recording data for each web conference. In the recording data storage unit 141, each piece of recording data is associated with the type of topic in the web conference represented by the recording data. The attention level calculation unit 204 calculates, for each piece of recording data, an attention level, which is an evaluation value based on the facial expression, actions, and speech content of at least one participant recorded in the recording data. Furthermore, when the reference information extraction unit 123 receives a designation of one piece of recorded data from the user via the input unit 161 or the communication unit 108, it acquires recorded data or conference information data that has the same topic type as the designated recorded data and has a difference in attention level within a predetermined range from the recorded data storage unit 141 or the conference information data storage unit 143, and presents information about the acquired recorded data or conference information data to the user via the output unit 162 or the communication unit 108. As a result, when making advance preparations for a subsequent web conference that has not yet been held after the web conference represented by the designated recorded data, the user can easily achieve the purpose of the web conference by appropriately referring to the contents of the recorded data or conference information data extracted by the reference information extraction unit 123. This allows the web conference support system 100 to provide web conference support that contributes to achieving the purpose of the web conference.

In the recording data storage unit 141 or the conference information data storage unit 143, the recording data or conference information data acquired by the reference information extraction unit 123 may be linked to the recording data or conference information data of at least one subsequent web conference among the one or more subsequent web conferences that are in the same chronological order as the recording data or conference information data and are one or more consecutive web conferences held after the web conference represented by the recording data or conference information data. In this case, the information displayed regarding the recording data or conference information data acquired by the reference information extraction unit 123 may include information regarding the recording data or conference information data of at least one subsequent web conference. In this case, the user can more easily achieve the purpose of the web conference by making advance preparations for the unheld web conference subsequent to the web conference represented by the recording data specified by the user, by referring to the information regarding the recording data or conference information data of the subsequent web conference.

The web conference support system 100 may include a conference information extraction unit 122. The conference information extraction unit 122 may generate conference information data for each piece of recorded data, the conference information data being data including a topic type linked to the recorded data and an attention level calculated for the recorded data, and store the conference information data linked to the recorded data in the conference information data storage unit 143. Information displayed regarding the recorded data acquired by the reference information extraction unit 123 may include information represented by the conference information data linked to the recorded data.
The conference information data is data that represents the details of the corresponding web conference and is presented to the user as text information. This allows the user to quickly and easily prepare for a subsequent web conference that has not yet been held after the web conference represented by the recording data or conference information data that the user has specified, making it easier to achieve the purpose of the web conference. This allows the web conference support system 100 to provide web conference support that contributes to achieving the purpose of the web conference.

For each piece of recorded data, the conference information data linked to the recorded data may further include time information indicating the conference duration from the start time to the end time of the web conference represented by the recorded data. An attention level may be linked to each point in the conference duration. Based on the conference information data linked to the acquired recorded data, the reference information extraction unit 123 may present to the user via the output unit 162 or the communication unit 108 a time series chart consisting of a first axis measuring time and a second axis measuring the attention level value.

Figure 16 shows an example of a timeline display screen.

The time series chart illustrated in FIG. 16 shows the correspondence between time and attention level. According to the display example illustrated in FIG. 16, the user can easily visually confirm the magnitude and degree of attention level for each hour. Furthermore, according to this configuration, as illustrated in FIG. 16, information such as the content of the speaker's utterance and the customer's emotions and actions at the time when an increase or decrease in attention level is observed is displayed on the time series chart, so that the user can easily and intuitively grasp the key points of the web conference represented by the conference information data. This allows the user to carry out more accurate advance preparations, and the web conference support system 100 can provide web conference support that contributes to achieving the purpose of the web conference.

For each piece of recorded data, the conference information data linked to that recorded data may further include information indicating the type of participants recorded in that recorded data. The reference information extraction unit 123 may present the information indicating the type of participants recorded in that recorded data to the user via the output unit 162 or the communication unit 108 as information related to the acquired recorded data. This allows the user to carry out more accurate advance preparations based on the tendencies of customers who are participants in the web conference.

For each of the recorded data, the conference information data associated with the recorded data is at least one of the following (A) and (B):
(A) a word that, when used in a subsequent web conference that is in the same time series as the web conference represented by the recorded data and is held after the web conference, is likely to result in a higher attention level being calculated for the recorded data of the subsequent web conference than if the word was not used in the subsequent web conference;
(B) a word that, when used in a subsequent web conference that is in the same chronological order as the web conference represented by the recorded data and held after the web conference, is likely to result in a lower attention level being calculated for the recorded data of the subsequent web conference than if the word was not used in the subsequent web conference;
It may further include.

Figure 17 shows an example of a screen displaying important/bad words.

In the display example shown in FIG. 17, words that contributed to increasing or decreasing the level of attention among those used during the web conference represented by the conference information data presented as reference information are displayed in a font size according to their frequency. In addition, important words that contributed to increasing the level of attention and bad words that contributed to decreasing the level of attention are displayed in different colors. These words extracted by the important/bad word extraction unit 205 are likely to become important words or bad words in subsequent web conferences that have not yet been held after the web conference represented by the recorded data or conference information data specified by the user. Therefore, with this configuration, the user can know in advance which words are likely to become important words or bad words in future web conferences, and can carry out advance preparations with this in mind.

For each of the recorded data, the conference information data linked to the recorded data may further include important material information, which is information indicating a material that is in the same time series as the web conference represented by the recorded data and is likely to result in a higher level of attention being calculated for the recorded data of the subsequent web conference when the important material information is used in the subsequent web conference, which is a web conference held after the recorded data, than when the important material information is not used.

Figure 18 shows an example of an important documents display screen.

In the display example shown in FIG. 18, materials that contributed to increasing or decreasing the level of attention among materials used during a web conference represented by the conference information data presented as reference information are displayed in association with the viewing time and the level of attention. These materials extracted by the used material extraction unit 206 are likely to be as useful as the materials used in a subsequent web conference that has not yet been held after the web conference represented by the recording data or conference information data specified by the user. Therefore, with this configuration, the user can know in advance which materials will be useful in future web conferences and can make advance preparations with this in mind.

The web conference support system 100 may include a recommendation target extraction unit 124. For each piece of recorded data, the conference information data linked to the recorded data may further include proposed target information, which is information related to a proposed target in the web conference represented by the recorded data. As an example, the proposed target may be a product or service to be proposed to a customer in a web conference for business negotiations. The recommendation target extraction unit 124 may extract a recommended target, which is a proposed target that is likely to have appeal to customers of the web conference, based on the conference information data, and the reference information extraction unit 123 may display information related to the extracted recommended target.

Figure 19 shows an example of the flow of the recommendation target extraction process.

The recommendation target extraction unit 124 refers to the conference information data stored in the conference information data storage unit 143 and identifies customers of other web conferences represented by the recording data or conference information data whose difference points are lower than a predetermined reference value (S1901). The customers identified in the process of S1901 are referred to as other customers. Next, the recommendation target extraction unit 124 acquires customer information data of the other customers from the customer information data storage unit 145 (S1902). Note that the customer information data may include data representing information regarding the customer's purchase history. In this case, the recommendation target extraction unit 124 extracts products or services included in the purchase history of the other customers as recommendation targets based on the acquired customer information data (S1903).

Next, the recommendation target extraction unit 124 calculates the reference purchase age by identifying the age at which other customers purchased the product or service extracted as a recommendation target in the process of S1903 based on the customer information data acquired in the process of S1902 (S1904).

Next, the recommendation target extraction unit 124 calculates the reference purchase time by identifying the time when other customers purchased the product or service extracted as a recommendation target in the process of S1903 based on the customer information data acquired in the process of S1902 (S1905).

Next, the recommendation target extraction unit 124 calculates reference purchase events by identifying events that occurred to other customers around the time when the product or service extracted as a recommendation target in the process of S1903 was purchased based on the customer information data acquired in the process of S1902 (S1906).

The recommendation target extraction unit 124 presents information about the product or service extracted as a recommendation target in the process of S1903, and information about the reference purchase age, reference purchase time, and reference purchase event calculated in each process of S1904 to S1906, to the user as recommendation information via the output unit 162 or the communication unit 108 (S1907), and the process ends.

As a result of the recommendation target extraction process, recommended targets for the customers the user is responsible for are extracted. In the recommendation target extraction process, other customers with similar tendencies to the customer the user is responsible for are identified based on the low difference points between the two recorded data or conference information data being compared. Therefore, products or services included in the purchase history of other customers are likely to be appealing to the customer the user is responsible for. In other words, a user of the web conference support system 100 can propose products or services that are likely to be appealing to the customer in question, for example, in a business negotiation that is the purpose of a web conference. As a result, the web conference support system 100 can provide web conference support that contributes to achieving the purpose of the web conference.

Although one embodiment has been described above, this is merely an example for the purpose of explaining the present invention, and is not intended to limit the scope of the present invention to only this embodiment. The present invention can also be implemented in various other forms.

100: Web conference support system

Claims (10)

A storage device;
a processor connected to the storage device;
The storage device stores recording data for each web conference,
A type of topic of the web conference represented by the recorded data is associated with each recorded data in the storage device;
The processor,
Calculating an attention level for each of the video recording data, which is an evaluation value based on the facial expression, the action, and the speech content of at least one participant recorded in the video recording data;
When a recording data is specified by the user,
acquiring from the storage device recorded data that has the same topic type as the designated recorded data and has a difference in attention level with respect to the designated recorded data within a predetermined range;
A web conferencing support system that displays information about acquired recording data. The web conference support system of claim 1, wherein when the acquired recording data is linked to recording data of at least one of one or more subsequent web conferences which are in the same chronological order as the recorded data and which are one or more consecutive web conferences held after the web conference represented by the recorded data, the information displayed regarding the acquired recording data includes information regarding the recording data of the at least one subsequent web conference. The processor, for each of the recording data,
generating meeting information data including a topic type associated with the recording data and a calculated level of attention for the recording data;
storing the conference information data linked to the recorded data in the storage device;
The web conference support system according to claim 2 , wherein the information displayed regarding the acquired recorded data includes information represented by conference information data linked to the recorded data. For each recording data,
the conference information data associated with the video recording data further includes time information representing a conference duration from a start time to an end time of the web conference represented by the video recording data;
The level of attention is associated with each point in time during the meeting.
The web conference support system of claim 3, wherein the processor displays a time series chart having a first axis measuring time and a second axis measuring attention value based on the conference information data linked to the acquired recording data. For each of the video recording data, the conference information data associated with the video recording data further includes information indicating a type of the participants recorded in the video recording data;
The web conference support system according to claim 3 , wherein the processor displays information indicating a type of the participants recorded in the acquired recorded data as the information related to the acquired recorded data. For each of the video recording data, the conference information data associated with the video recording data further includes at least one of the following (A) and (B):
(A) a word that, when used in a subsequent web conference that is in the same time series as the web conference represented by the recorded data and is held after the web conference, is likely to result in a higher attention level being calculated for the recorded data of the subsequent web conference than if the word was not used in the subsequent web conference;
(B) a word that, when used in a subsequent web conference that is in the same chronological order as the web conference represented by the recorded data and held after the web conference, is likely to result in a lower attention level being calculated for the recorded data of the subsequent web conference than if the word was not used in the subsequent web conference;
The web conference support system according to claim 3 . The web conference support system of claim 3 further includes important material information, which is information indicating materials that, for each recorded data, are in the same chronological order as the web conference represented by the recorded data, and that, if used in a subsequent web conference that is a web conference held after the current web conference, are likely to result in a higher level of attention being calculated for the recorded data of the subsequent web conference than if the materials were not used. For each recording data,
The conference information data associated with the recording data further includes proposed target information that is information regarding a proposed target in the web conference represented by the recording data;
The processor extracts, based on the conference information data, a recommendation target that is a proposal target that is likely to be appealing to participants of the web conference;
The web conference support system according to claim 3 , further comprising: displaying information relating to the extracted recommendation targets. The computer stores the recording data for each web conference,
Each piece of recorded data is associated with a type of topic of the web conference represented by the recorded data and stored;
The computer calculates, for each of the video recording data, an attention level which is an evaluation value based on the facial expression, the action, and the speech content of at least one participant recorded in the video recording data;
When the computer receives a request for a recording from the user,
extracting from the data set of the stored recording data recording data, recording data having the same topic type as the designated recording data and a difference in attention level with respect to the designated recording data within a predetermined range;
A method for supporting a web conference that displays information about extracted recorded data. Stores recording data for each web conference,
Each piece of recorded data is associated with a type of topic of the web conference represented by the recorded data and stored;
Calculating an attention level for each of the video recording data, which is an evaluation value based on the facial expression, the action, and the speech content of at least one participant recorded in the video recording data;
When a recording data is specified by the user,
extracting from the data set of the stored recording data recording data, recording data having the same topic type as the designated recording data and a difference in attention level with respect to the designated recording data within a predetermined range;
A computer program product that causes a computer to display information about the extracted recorded data.

Images