JP2019021337A - Conference support device, conference support device control method, and program - Google Patents

Abstract

To provide a conference support device for facilitating presentation of a topic to a user, that is the topic for developing a discussion such as a related topic that is not discussed in a conference.SOLUTION: A conference support device includes: acquisition means for acquiring minute data of a conference on the basis of voice data or text data received from one or more terminal devices; recording means for recording the acquired minute data in a minute database for storing a plurality of pieces of minute information; topic extraction means for extracting a plurality of topics on the basis of the minute database and an external database different from the minute database; keyword extraction means for extracting a keyword related to the conference on the basis of the acquired minute data; and display control means for displaying the information of the topic related to the extracted keyword among the plurality of the extracted topics.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a conference support apparatus, a control method for a conference support apparatus, and a program.

  In order to facilitate a meeting, whether the content currently being spoken matches the goal of the meeting, and if similar meetings have been held in the past, the contents developed Or whether similar discussions are repeated.

  As a device for supporting such a conference, a related conference document is acquired, the importance of each phrase is calculated using the reverse frequency for each conference document, and a predetermined number of keywords are determined based on the calculated importance. A conference keyword extracting device for extracting is known (for example, see Patent Document 1).

  Since the technique disclosed in Patent Document 1 extracts keywords based on meeting documents such as minutes related to a meeting, for example, a topic such as a topic that has not yet been discussed in a meeting, such as a topic that develops the discussion. It was difficult to present.

  The embodiment of the present invention has been made in view of the above problems, and conference support that makes it easy to present to the user topics that develop discussion, such as topics that have not yet been discussed in the conference. An object is to provide an apparatus.

  In order to solve the above problems, a conference support device according to an embodiment of the present invention includes an acquisition unit that acquires conference proceedings data based on voice data or text data received from one or more terminal devices, Recording means for recording the acquired minutes data in a minutes database for storing minutes information, topic extraction means for extracting a plurality of topics based on the minutes database and an external database different from the minutes database And a keyword extracting means for extracting a keyword related to the meeting based on the acquired agenda data; a display control means for displaying information on a topic related to the extracted keyword among the extracted plurality of topics; Have

  According to the embodiment of the present invention, it is possible to provide a conference support apparatus that facilitates presenting a user with a topic in which discussion develops, such as a topic not yet discussed in a conference.

It is a figure showing an example of composition of an information processing system concerning one embodiment. It is a figure which shows an example of the hardware constitutions of the meeting assistance apparatus which concerns on one Embodiment. It is a functional block diagram of the meeting assistance apparatus which concerns on one Embodiment. It is a figure which shows the example of the display item of the display screen which concerns on one Embodiment. It is a figure which shows the example of a display of the display screen which concerns on one Embodiment. It is a flowchart which shows the flow of a process of the meeting assistance apparatus which concerns on one Embodiment. It is a flowchart which shows the flow of the process regarding the external data which concerns on one Embodiment. It is a figure which shows the example of the information memorize | stored in the external data database which concerns on one Embodiment. It is a flowchart which shows the flow of the process regarding the minutes data which concerns on one Embodiment. It is a figure which shows the example of the information memorize | stored in the minutes database which concerns on one Embodiment.

  Embodiments of the present invention will be described below with reference to the accompanying drawings.

<System configuration>
It is a figure showing an example of composition of an information processing system concerning one embodiment. An information processing system (conference system) 100 includes a conference support apparatus 101, a display apparatus 102, participant terminals 103 and 104, an organizer terminal 105, and the like. In addition, the conference support apparatus 101, the participant terminals 103 and 104, and the organizer terminal 105 are connected to a network such as the Internet or a wireless / wired LAN (Local Area Network) and can perform data communication with each other. To do.

  The conference support apparatus 101 is an information processing apparatus such as a server. Each function according to the present invention is realized by, for example, a program executed by the conference support apparatus 101.

  The display device 102 is a display device such as a projector or an electronic blackboard. The organizer terminal 105 is an information device such as a notebook PC (Personal Computer), for example, and is connected to the display device 102 and displays a display screen on the display device 102 in the example of FIG. The participant terminals 103 and 104 are information devices such as notebook PCs, tablet terminals, and smartphones, for example. The participant terminals 103 and 104 and the organizer terminal 105 are assumed to have a microphone for voice input.

  In FIG. 1, a moderator 107 who holds a conference 106 accesses the conference support apparatus 101 via the network using the organizer terminal 105 and performs an operation for starting the conference. The conference support apparatus 101 performs a conference 106 start process in response to a conference start operation.

  The participants 108 and 109 of the conference 106 can access the conference support apparatus 101 using the participant terminals 103 and 104, respectively, and can participate in the conference 106 held by the chairperson 107 via the network.

  When the conference 106 starts, the conference support apparatus 101 distributes display data such as a shared screen, synchronization information, and the like to the organizer terminal 105 and the participant terminals 103 and 104. The display device 102, the organizer terminal 105, the participant terminals 103 and 104, etc. display the conference screen of the conference 106 based on the distributed data, and also to the display device 102 or the organizer terminal 105 of the moderator 107. This operation is also reflected on the display screens of the participant terminals 103 and 104, for example.

  With the above configuration, the moderator 107 and the participants 108 and 109 can hold a conference while sharing information on the shared screen distributed from the conference support apparatus 101.

  Note that the configuration of the information processing system 100 in FIG. 1 is merely an example, and does not limit the scope of the present invention. For example, the numbers of the conference support apparatus 101, the display apparatus 102, the participant terminals 103 and 103, the organizer terminal 105, and the like are examples, and other numbers may be used. Further, the function of the conference support apparatus 101 may be divided and provided on a plurality of servers.

<Hardware configuration>
FIG. 2 is a diagram illustrating an example of a hardware configuration of the conference support apparatus according to the embodiment. The conference support apparatus 101 has a general computer configuration. For example, a CPU (Central Processing Unit) 201, a RAM (Random Access Memory) 202, a ROM (Read Only Memory) 203, a storage unit 204, an external I / O. / F (Interface) unit 205, input unit 206, display unit 207, communication I / F unit 208, bus 209, and the like.

  The CPU 201 is an arithmetic device that implements each function of the computer 200 by reading programs and data from the ROM 203, the storage unit 204, and the like onto the RAM 202 and executing the processing. The RAM 202 is a volatile memory used as a work area for the CPU 201. The ROM 203 is a non-volatile memory that can retain programs and data even when the power is turned off. The ROM 203 stores, for example, BIOS (Basic Input Output System) executed when the computer 200 is started up, OS (Operating System) settings, and the like.

  The storage unit 204 is a storage device (storage device) such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive), and stores an OS, application programs, and various data. The external I / F unit 205 is an interface with an external device. The external device includes a recording medium 210, for example. The conference support apparatus 101 can read and / or write the recording medium 210 via the external I / F unit 205. The recording medium 210 includes, for example, an optical disk, a magnetic disk, a memory card, a USB (Universal Serial Bus) memory, and the like.

  The input unit 206 is a device that includes a pointing device such as a mouse, a keyboard, and the like, and inputs operation signals to the computer 200. The display unit 207 includes a display and the like, and displays the processing result of the conference support apparatus 101 and the like. The communication I / F unit 208 is an interface that connects the conference support apparatus 101 to a network. The conference support apparatus 101 can transmit and receive data to and from other information devices via the communication I / F unit 208. A bus 209 is connected to each of the above components, and transmits an address signal, a data signal, various control signals, and the like.

  Each function of the conference support apparatus 101 is realized by a program executed by the CPU 201, for example. 2 is an example of the hardware configuration of the conference support apparatus 101, and does not limit the scope of the present invention. For example, the conference support apparatus 101 may include the input unit 206 and the display unit 207 outside, or may not necessarily include the input unit 206 and the display unit 207.

  The participant terminals 103 and 104 and the organizer terminal 105 have, for example, a general computer configuration as shown in FIG. 2, and further have a microphone for acquiring audio data. It shall be.

<Functional configuration>
FIG. 3 is a functional configuration diagram of the conference support apparatus according to the embodiment. 3, the conference support apparatus 101 includes, for example, an OS 301, middleware 302, a communication unit 303, an engine unit 304, a minutes DB (Database) 305, an external data DB 306, a user information DB 307, a dictionary DB 308, and the like.

  The OS 301 is an operating system such as Linux (registered trademark). The middleware 302 is a library group that operates on the OS 301 and is used by the communication unit 303, the engine unit 304, and the like.

  The communication unit 303 is a communication unit for connecting the conference support apparatus 101 to a network and transmitting / receiving data. The communication unit 303 includes, for example, the communication I / F unit 208 of FIG. 2 and its control unit. The communication unit 303 transmits / receives data to / from the participant terminals 103 and 104, the organizer terminal 105, and the like using a communication protocol such as HTTP (Hyper Text Transfer Protocol).

  The engine unit 304 includes a voice data acquisition unit 310, a voice recognition unit 311, a text data acquisition unit 312, a minutes creation unit 313, a topic extraction unit 314, an external data acquisition unit 315, a morpheme analysis unit 316, a topic search unit 317, a topic A display unit 318, a minutes display unit 319, a mind map generation unit 320, a mind map display unit 321 and the like are included.

  The audio data acquisition unit 310 acquires audio data transmitted from the participant terminals 103 and 104 participating in the conference 106, the organizer terminal 105, and the like via the communication unit 303. The participant terminals 103 and 104 and the organizer terminal 105 convert the audio signal collected by the microphone during the conference into predetermined audio data and transmit it to the conference support apparatus 101.

  The voice recognition unit 311 performs voice recognition processing on the voice data acquired by the voice data acquisition unit 310, and converts voice data of the moderator 107, participants 108, 109, etc. in the conference 106 into text data. The speech recognition unit 311 performs speech recognition processing using, for example, an acoustic model stored in the dictionary DB 308.

  The text data acquisition unit 312 acquires text data input from the participant terminals 103 and 104, the organizer terminal 105, and the like via the communication unit 303.

  The voice data acquisition unit 310, the voice recognition unit 311, and the text data acquisition unit 312 acquire the agenda data 309 for the conference 106 based on voice data or text data received from one or more terminal devices. It is an example.

  The minutes creation unit 313 is a recording unit that records the minutes data such as text data acquired by the text data acquisition unit 312 and / or text data converted into text by the voice recognition unit 311 in the minutes DB 305.

  The topic extraction unit 314 is a topic extraction unit that extracts a plurality of topics (topics) based on the minutes DB 305 and the external data DB 306. Note that a topic is a word (keyword) or a collection of words that represents the sentence.

  In the present embodiment, it is assumed that the topic extraction unit 314 extracts topics using a general topic model. For example, the topic extraction unit 314 can extract a topic by TF-IDF (Term Frequency-Inverse Document Frequency) as in Patent Document 1. The TF-IDF breaks a sentence into morphemes, calculates an index indicating the importance such as IF (word appearance frequency) -IDF (reverse appearance frequency) value for each morpheme, and selects a morpheme having a high index. It is the technology that is the topic of the sentence. The TF-IDF has a feature that it is easy to implement because the topic extraction algorithm is relatively simple.

  For example, the topic extraction unit 314 acquires the data of the minutes DB 305 and the external data DB 306, extracts topics from the acquired data, and records the extracted topics in the minutes DB 305 and the external data DB 306.

  The external data acquisition unit 315 downloads a public database (external database) such as Wikipedia on the Internet via the communication unit 303 and stores it in the external data DB 306. Wikipedia on the Internet is a suitable example of an external database, and the external database includes various databases different from the minutes DB 305, such as minutes databases of other departments that can be accessed on a corporate LAN. .

  The morpheme analysis unit 316 decomposes the agenda data such as text data acquired by the text data acquisition unit 312 or text data converted into text by the speech recognition unit 311 into parts of speech level words (morphemes), and converts them into decomposed words. Add part-of-speech information. In addition, the morphological analysis unit 316 extracts, for example, nouns from the decomposed words as keywords related to the agenda content of the meeting 106. Note that the morphological analysis unit 316 is an example of a keyword extraction unit according to the present invention.

  The topic search unit 317 searches for a topic related to the keyword extracted by the morpheme analysis unit 316 among the plurality of topics extracted by the topic extraction unit 314. The topic display unit 318 performs control to display information on the topic searched by the topic search unit 317 on the display device 102, the organizer terminal 105, the participant terminals 103 and 104, and the like.

  Note that the topic search unit 317 and the topic display unit 318 are examples of display control means according to the present invention.

  The minutes display unit 319 displays link information to the minutes including the topics extracted from the minutes DB 305 among the topics searched by the topic search unit 317, the display device 102, the organizer terminal 105, and the participant terminals 103. , 104 and the like.

  For example, the mind map generation unit 320 receives an operation performed by the moderator 107 on the mind map displayed on the display device 102 via the communication unit 303, and updates the mind map by reflecting the operation result. . The mind map display unit 321 displays the mind map generated by the mind map generation unit 320 on the display device 102, the organizer terminal 105, the participant terminals 103 and 104, and the like.

  A mind map is a thinking tool that illustrates thought by placing keywords and the like that are the center of the concept to be expressed in the center, and then expanding and connecting the keywords, topics, and the like radially. In the mind map, newly added topics are connected by a curve, for example.

  The mind map generation unit 320 and the mind map display unit 321 are examples of the mind map display unit according to the present invention.

  The minutes DB 305 is a database that stores past meeting data (for example, minutes data, etc.) created by the meeting support apparatus 101. The minutes DB 305 also stores topic information of the minutes DB 305 extracted by the topic extraction unit 314.

  The external data DB 306 is a database that stores data acquired from an external database such as a public database of Wikipedia, for example. The external data DB 306 also stores topic information of the external data DB 306 extracted by the topic extraction unit 314.

  The user information DB 307 is a database for managing information on users who use the information processing system 100, for example, the moderator 107 and the participants 108 and 109. The dictionary DB 308 stores data such as an acoustic model used by the speech recognition unit 311 and a word model used by the morpheme analysis unit 316.

  Note that the configuration in FIG. 3 is an example, and the conference support apparatus 101 uses at least a part of the minutes DB 305, external data DB 306, user information DB 307, and dictionary DB 308 as an external storage device, external server, or database device. And so on.

<Example of display screen>
FIG. 4 is a diagram illustrating an example of display items on the display screen according to an embodiment. On the display screen 400 of the display device 102 or the like, for example, an agenda 401, a topic 402, agenda content 403, minutes topic information 404, external topic information 405, etc. are displayed.

  A conference agenda 401 displays words that are agenda for the conference.

  In the topic 402, the topic currently being discussed is displayed.

  In the agenda content 403, speech contents of the moderator 107 and the participants 108 and 109 converted into text by the conference support apparatus 101, input text, and the like are displayed in text. Furthermore, in the agenda content 403, words (keywords) that become topics such as nouns extracted by the morphological analysis unit 316 of the conference support apparatus 101 are underlined and displayed separately from other words.

  A plurality of minutes topics searched by the topic search unit 317 among a plurality of topics extracted from the minutes DB 305 by the topic extraction unit 314 are displayed in the upper part of the minutes topic information 404. In the lower part of the minutes topic information 404, link information to the minutes related to the plurality of minutes topics displayed in the upper part is displayed. By selecting this link information, for example, the minutes of the link destination can be displayed on the display screen 400.

  The external topic information 405 displays a plurality of external topics searched by the topic search unit 317 among a plurality of topics extracted by the topic extraction unit 314 from the external data DB 306.

  FIG. 5 is a diagram illustrating a display example of a display screen according to an embodiment. For example, when the conference presenter 107 starts the conference 106 using the organizer terminal 105 and inputs the conference agenda, the input agenda 401 is displayed near the center of the display screen 501 as shown in the display screen 501. The

  In the agenda content 403 of the display screen 501, the agenda content currently being discussed by the moderator 107, participants 108, 109, etc. of the conference 106 is received from the speech recognition unit 311 and the text data acquisition unit 312 of the conference support apparatus 101. Text is displayed based on the output data.

  The minutes topic information 404 on the display screen 501 displays a plurality of minutes topics searched by the topic search unit 317 and link information to the minutes including the minutes topics.

  A plurality of external topics searched by the topic search unit 317 are displayed in the external topic information 405 on the display screen 501.

  Next, for example, when the moderator 107 operates the organizer terminal 105 to select one topic from the external topic information 405 or inputs a new topic, the mind map generation unit 320 and the mind map display unit 321 As a result, the mind map is updated. For example, as shown in the display screen 502, the selected or input topic 402 is displayed together with the new node.

  Further, on the display screen 502, the agenda content 403 on the display screen 502 is updated according to the agenda content discussed by the presenter 107, the participants 108, 109, and the like of the conference 106. Furthermore, an underline is added to a keyword (for example, a noun) extracted by the morphological analysis unit 316 in the agenda content 403. In addition, when the keyword displayed on the agenda content 403 increases, for example, the keyword may be selectively displayed according to the appearance frequency. Alternatively, for example, a part of the keyword may be selectively displayed by filtering in a technical field or the like.

  The minutes topic information 404 on the display screen 502 includes a plurality of minutes topics newly searched by the topic search unit 317 according to the keywords extracted by the morphological analysis unit 316 and minutes to the minutes including the minutes topics. Link information and the like are displayed.

  In the external topic information 405 on the display screen 502, a plurality of external topics newly searched by the topic search unit 317 are displayed according to the keywords extracted by the morphological analysis unit 316.

  Further, when the moderator 107 operates the organizer terminal 105 to select one topic from the external topic information 405 or inputs a new topic, the mind map generation unit 320 and the mind map display unit 321 cause the mind to be displayed. The map is updated. For example, as shown in the display screen 503, the selected or input topic 402 is displayed together with the new node. Also, on the display screen 502, different topics are displayed as shown in the screen 504, for example, according to the topic selected or input from the external topic information 405 by the moderator 107.

  As described above, in the information processing system 100 according to the present embodiment, for example, topics related to the agenda contents such as the utterance contents of the moderator 107 and / or the participants 108 and 109 include the minutes topic information 404 and the external topics. Information 405 is sequentially presented to the user.

  Further, the minutes topic information 404 displays topics related to the contents discussed in the past meetings, and link information for accessing the minutes including the displayed topics. As a result, the user can easily determine whether the same meetings have been held in the past, whether the same discussion as in the past meetings has been repeated, or whether the contents have been developed with respect to the past discussion contents. Can be confirmed.

  Furthermore, since the external topic information 405 also displays topics that have not yet been discussed in past meetings, the user can be encouraged to discuss new topics.

<Process flow>
FIG. 6 is a flowchart illustrating a process flow of the conference support apparatus according to the embodiment. When the conference start operation is performed by the moderator 107 of the conference 106, the conference support apparatus 101 performs a conference start process (step S601). In the process of starting the meeting, for example, acquisition of a meeting ID, recording of information such as date and time, meeting organizer, and the like are performed, and recording of minutes data in the minutes DB 305 by the minutes creation unit 313 is also started.

  The voice data and text data transmitted from the participant terminals 103 and 104 and the organizer terminal 105, where the conference start process is performed, are received by the communication unit 303 and sent to the voice data acquisition unit 310 and the text data acquisition unit 312. Input (steps S602 and S610).

  The voice data acquired by the voice data acquisition unit 310 is subjected to voice recognition processing by the voice recognition unit 311 and converted into text data (step S603).

  Text data acquired by the text data acquisition unit 312 and text data (meeting data) converted by the voice recognition unit 311 are passed to the morphological analysis unit 316 and stored in the minutes DB 305 by the minutes creation unit 313 ( Step S611).

  The morpheme analysis unit 316 divides the input text data (procedure data) into morphemes (words) and adds each part of speech information to generate morpheme information (step S604). From the generated morpheme information, a morpheme to be extracted is selected based on the part of speech (step S605). In the present embodiment, nouns are selected (extracted) in step S605. Note that the process of selecting a noun in step S605 is an example. In step S605, for example, as described above, the keyword may be selected based on the appearance frequency or the like, or the keyword may be selected by filtering under a specific condition such as a technical field.

  The morpheme analysis unit 316 passes the noun (keyword) extracted in steps S604 and S605 to the topic search unit 317.

  The topic search unit 317 searches the topic stored in the minutes DB 305 for the same topic as the noun (keyword) extracted by the morphological analysis unit 316 (step S606). At this time, a plurality of extracted nouns may be used as search terms. In this case, a vector using the appearance frequency is created for a plurality of search terms, and the topic search unit 317 calculates the similarity based on this vector.

  The topic search unit 317 passes, from the search result, a topic group including the same topic as the keyword or a topic having a high similarity and a link of the minutes to the topic display unit 318.

  Similarly, the topic search unit 317 searches the topic stored in the external data DB 306 for the same topic as the noun (keyword) extracted by the morpheme analysis unit 316 (step S612). Also, the topic search unit 317 passes a topic group including the same topic as the keyword or a topic having a high similarity from the search result to the topic display unit 318.

  The topic display unit 318 displays the topic information related to the keyword searched by the topic search unit 317 on the display device 102, the participant terminals 103, 104, and the like. Specifically, the topic group searched from the minutes DB 305 and the link of the minutes are displayed as the minutes topic information 404 in FIG. The topic group retrieved from the external data DB 306 is displayed as the external topic information 405 in FIG.

  For example, the topic display unit 318 generates the display screen 400 including the minutes topic information 404 and the external topic information 405, and transmits the display screen 400 to the display device 102 and the participant terminals 103 and 104 for display. Alternatively, the topic display unit 318 may transmit the above-described minutes topic information 404 and external topic information 405 to the display device 102 and the participant terminals 103 and 104, and cause the destination terminal to create a display screen. good.

  If the conference continues in step S608, the conference support apparatus 101 returns to step S602 and step S610 and repeats the above processing. On the other hand, if the meeting is not continued in step S608, processing for extracting a topic from the minutes data is performed (step S609).

<External data DB>
Since it takes time to extract a topic from an external database such as a public database on the Internet, it is desirable to perform the process separately from the conference and store the topic in the external data DB 306 in advance.

  FIG. 7 is a flowchart showing a flow of processing regarding external data according to an embodiment. The external data acquisition unit 315 acquires data of a public database such as Wikipedia on the Internet, for example, via the communication unit 303 (step S701) and stores it in the external data DB 306 (step S702).

  Subsequently, in step S703 and subsequent steps, processing for extracting a topic from the database stored in the external data DB 306 is performed. This processing may be performed after step S702, or may be performed separately. good.

  In step S703, the conference support apparatus 101 acquires public database data stored in the external data DB 306, and selects data necessary for topic extraction (step S704). In this embodiment, a title, ID, summary, or text is selected as data necessary for topic extraction. In addition, morphological analysis is performed on the selected summary and text (step S705).

  Further, the conference support apparatus 101 extracts a noun from the morphological-analyzed data, creates a corpus by creating a frequency matrix (step S706), and extracts a topic based on the corpus (step S707). . In the present embodiment, for example, the above-described TF-IDF is used as a topic extraction method.

  Further, the conference support apparatus 101 stores the extracted topic in the external data DB 306 for each title (step S708).

  FIG. 8 is a diagram illustrating an example of information stored in an external data database according to an embodiment. The data 801 in FIG. 8 is an example of data acquired from a public database such as Wikipedia, and is described in XML (Extensible Markup Language). In the example of FIG. 8, the data 801 includes information such as a title 802, an ID 803 that identifies the title 802, and a text 804 that describes the title 802.

  The conference support apparatus 101 stores the acquired data 801 as it is in the external data DB 306 in steps S701 and S702 of FIG. In step S703 in FIG. 7, data is acquired in the format of data 801. In step S704, information on the <id> tag and <text> tag in the <page> tag is extracted from the data 801, for example, Then, data as shown in the table information 805 is generated. In the table information 805 in FIG. 8, an ID 803 (page-id) and a text 804 (text) are stored in association with each other.

  Further, in step S708 in FIG. 7, the conference support apparatus 101 creates the data extracted as shown in the table information 806 in FIG. 8 for the topic extracted in step S707, for example. In FIG. 8, in the table information 806, an ID 803 (page-id) and a plurality of topics (for example, nouns) are stored in association with each other. The conference support apparatus 101 stores the created table information 805, 806, etc. in the external data DB 306.

<Meeting DB>
The process of extracting topics from the minutes data is performed, for example, after the conference ends (step S609 in FIG. 6).

  FIG. 9 is a flowchart showing a flow of processing regarding minutes data according to an embodiment. When the conference ends, the conference support apparatus 101 acquires the minutes data of the ended conference from the minutes DB 305 (step S901), and selects necessary data, for example, a title and a text from the acquired data (step S902). . Also, morphological analysis is performed on the selected text (step S903), and nouns are extracted. Further, a corpus is generated from the extracted nouns by creating a frequency matrix (step S904), and topics are extracted based on the corpus (step S905). Further, the conference support apparatus 101 stores the extracted topic in the minutes DB 305 for each minutes (step S906).

  FIG. 10 is a diagram illustrating an example of information recorded in the minutes database according to the embodiment. The minutes data 1001 created by the minutes creation unit 313 includes, for example, “ID”, “date and time”, “creation ID”, “title”, “text”, etc., as shown in FIG.

  In the minutes data 1001, “ID” is identification information for specifying a meeting or the like. “Date and time” is the date and time when the minutes were created, and may be the date and time when the creation of the minutes or the time when the created minutes were recorded, for example. “Title” is the title of the agenda such as a meeting. “Body” includes agenda data, for example, text data acquired from the speech recognition unit 311, the text data acquisition unit 312, and the like. The minutes data 1001 is described in, for example, SQL (Structured Query Language).

  In step S901 in FIG. 9, the conference support apparatus 101 acquires data in the format of minutes data 1001, and performs a series of processes from step S902 to S905. In step S708, the topic extracted in step S905 is created, for example, as shown in the table information 1002 in FIG. In FIG. 10, in the table information 1002, “ID” for specifying a meeting or the like and “topic” which is a plurality of minutes topics are stored in association with each other. The conference support apparatus 101 stores the created table information 1002 in the minutes DB 305.

  The conference support apparatus 101 can extract the minutes including a specific topic based on the minutes data 1001 and the “ID” of the table information 1002.

[Other Embodiments]
In the above embodiment, the topic extraction unit 314 has been described as performing topic extraction using TF-IDF. However, the topic extraction unit 314 extracts topics using other general topic models. It may be what performs.

  For example, LDA (Latent Dirichlet Allocation) is known as an example of another general topic model (see, for example, Non-Patent Document 1). The topic extraction unit 314 performs topic extraction using LDA, and the topic search unit 317 performs keyword search by calculating the similarity, so that even if there is no word that matches the search word in the sentence, it looks similar You will be able to extract words that are used in different ways.

  As another example of another general topic model, DTM (Dynamic Topic Model) is known (see, for example, Non-Patent Document 2). The topic extraction unit 314 performs topic extraction using DTM, and the topic search unit 317 performs keyword search by calculating similarity, so that changes in topics in a plurality of minutes arranged in time series can be extracted. become. Thereby, for example, topics that span a plurality of minutes can be extracted.

<Summary>
As described above, the conference support apparatus 101 according to the present invention acquires the agenda data of the conference 106 based on the voice data or the text data received from one or more terminal devices (the organizer terminal 105 and the participant terminals 103 and 104). An acquisition unit 309 is included. In addition, the conference support apparatus 101 includes recording means (minutes creation unit 313) for recording the obtained minutes data in a minutes database (minutes DB 305) that stores a plurality of minutes information. Furthermore, the conference support apparatus 101 includes topic extraction means (topic extraction unit 314) that extracts a plurality of topics based on the minutes database and an external database (external data DB 306 or the like) different from the minutes database.

  The conference support apparatus 101 also includes keyword extraction means (topic search unit 317) that extracts keywords related to the conference 106 based on the acquired agenda data. Further, the conference support apparatus 101 includes display control means (topic search unit 317, topic display unit 318) that displays topic information related to the extracted keyword among the plurality of extracted topics.

  With the above configuration, in the information processing system 100 according to the present embodiment, for example, topics related to the agenda contents such as the utterance contents of the moderator 107 and the participants 108 and 109 include the minutes topic information 404 and the external topic information. It is presented to the user as 405 one after another. At this time, topics not yet discussed in the past meetings are displayed in the external topic information 405, so that the user can be encouraged to discuss on a new issue.

  That is, according to the present embodiment, it is easy to display a topic on which the discussion develops, such as a topic not yet discussed in the conference, on the display device 102, the organizer terminal 105, the participant terminals 103, 104, and the like. The conference support apparatus 101 can be provided.

  Further, the minutes topic information 404 displays topics related to the contents discussed in the past meetings, and link information for accessing the minutes including the displayed topics. As a result, the user can easily determine whether the same meetings have been held in the past, whether the same discussion as in the past meetings has been repeated, or whether the contents have been developed with respect to the past discussion contents. Can be confirmed.

100 Information Processing System 101 Conference Support Device 103, 104 Participant Terminal (Terminal Device)
105 Organizer terminal (terminal equipment)
309 Acquisition means 313 Minutes creation section (recording means)
314 Topic extraction unit (topic extraction means)
316 Morphological analyzer (keyword extraction means)
317 Topic Search Unit 318 Topic Display Unit (Display Control Unit)

JP 2012-38064 A

D. M. Blei, A. N. Ng, and M. I. Jordan. Latent dirichlet allocation. In Journal of Machine Learning Research archive., 2003. D. M. Blei and J. D. Lafferty. Dynamic Topic Models. In Proc. Of the 23rd international conference on Machine learning., 2006.

In order to solve the above problem, a conference support device according to an embodiment of the present invention includes an acquisition unit that acquires conference proceedings data based on voice data or text data received from one or more terminal devices, and the acquisition from the proceedings data unit is acquired, and the keyword extracting means for extracting a keyword related proceedings content of the conference, other different database than the previous conference proceedings information stored minutes databases, and the minutes database From the topic extraction means for extracting the topic related to the keyword, the topic related to the keyword extracted from the minutes database according to the content of the meeting, the topic related to the keyword extracted from the other database, Display control means for displaying on the display screen .

Claims (10)

Acquisition means for acquiring conference proceedings data based on voice data or text data received from one or more terminal devices;
Recording means for recording the acquired minutes data in a minutes database storing a plurality of minutes information;
Topic extraction means for extracting a plurality of topics based on the minutes database and an external database different from the minutes database;
Keyword extracting means for extracting keywords related to the meeting based on the acquired agenda data;
Display control means for displaying information on a topic related to the extracted keyword among the plurality of extracted topics;
A conference support apparatus.   The conference support apparatus according to claim 1, wherein the external database is a public database on the Internet.   The meeting support apparatus according to claim 1, further comprising: minutes display means for displaying link information to minutes including a topic extracted from the minutes database among topics related to the extracted keyword.   4. The conference support apparatus according to claim 1, further comprising a mind map display unit configured to display a topic selected from the displayed topic information as a mind map. 5.   The conference support apparatus according to claim 1, wherein the acquisition unit includes a voice recognition unit that converts the voice data into text data.   The conference support apparatus according to claim 1, wherein the topic extraction unit extracts the plurality of topics by TF-IDF. The topic extraction means extracts the plurality of topics by LDA,
The conference support apparatus according to claim 1, wherein the keyword extraction unit extracts the keyword based on a similarity calculation. The topic extraction means extracts the plurality of topics by DTM,
The conference support apparatus according to claim 1, wherein the keyword extraction unit extracts the keyword based on a similarity calculation. Obtaining conference proceedings data based on voice data or text data received from one or more terminal devices;
Recording the acquired minutes data in a minutes database storing a plurality of minutes information;
Extracting a plurality of topics based on the minutes database and an external database different from the minutes database;
Extracting keywords related to the meeting based on the acquired agenda data;
Displaying information on a topic related to the extracted keyword among the plurality of extracted topics;
Control method for conference support apparatus including Acquisition means for acquiring conference proceedings data based on voice data or text data received from one or more terminal devices;
Recording means for recording the acquired minutes data in a minutes database storing a plurality of minutes information;
Topic extraction means for extracting a plurality of topics based on the minutes database and an external database different from the minutes database;
Keyword extracting means for extracting keywords related to the meeting based on the acquired agenda data;
Display control means for displaying information on a topic related to the extracted keyword among the plurality of extracted topics;
As a program to make the computer function as.

Images