Information processing system, program, and information processing method
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2026-01-09
- Publication Date
- 2026-08-13
Smart Images
Figure JP2026000540_13082026_PF_FP_ABST
Abstract
Description
Information Processing System, Program, and Information Processing Method
[0001] The present invention relates to an information processing system, a program, and an information processing method.
[0002] Conventionally, a technology for automatically generating minutes of a meeting has been known. Patent Document 1 discloses a system that detects a change in a topic based on attached information of display data and creates topic information for each topic.
[0003] Japanese Patent Application Laid-Open No. 2015-232769
[0004] However, it is desired to provide summary information that is easier to understand the content of the speech.
[0005] The present invention has been made in consideration of such points, and an object thereof is to provide an information processing system, a program, and an information processing method capable of presenting summary information that is easy to understand the content of the speech.
[0006] The information processing system of the present invention is an information processing system including an acquisition unit and an interaction unit, wherein the acquisition unit acquires speech information of a user, and the interaction unit includes a summary generation request for generating summary information including partial summary information of the speech information in a predetermined processing section each time the speech information in the processing section is obtained, and a prompt including an editing request for editing the already generated summary information based on the newly obtained partial summary information of the processing section. By inputting the speech information into a generation model, first summary information including partial summary information of a first processing section is obtained. After obtaining the first summary information, second summary information including partial summary information of the first processing section and partial summary information of a second processing section later than the first processing section is obtained. The partial summary information of the first processing section included in the second summary information is edited based on the partial summary information of the second processing section.
[0007] The information processing system of the present invention includes an editing request in which the prompt modifies the generated summary information based on newly generated partial summary information, and the second summary information may include partial summary information of the first processing section modified based on the partial summary information of the second processing section.
[0008] The information processing system of the present invention includes an editing request in which the prompt deletes the generated partial summary information based on the newly generated partial summary information, and the second summary information does not necessarily include the partial summary information of the first processing interval.
[0009] The information processing system of the present invention includes a prompt request that, when the partial summary information is deleted, include information indicating that the partial summary information has been deleted in the summary information, and the second summary information may include information indicating that the partial summary information of the first processing interval has been deleted.
[0010] In the information processing system of the present invention, if the user requests the display of the deleted partial summary information in the second summary information, the dialogue unit may input a prompt to the generation model that includes a request to add the deleted partial summary information, thereby obtaining the deleted partial summary information of the first processing section.
[0011] The information processing system of the present invention includes an editing request in which the prompt does not add the newly generated partial summary information to the summary information if the content of the newly generated partial summary information differs from the content of the previously generated summary information, and if the content of the partial summary information for a third processing section after the second processing section differs from the content of the second summary information, the third summary information generated after the second summary information may not include the partial summary information for the third processing section.
[0012] The information processing system of the present invention includes an editing request in which the prompt integrates newly generated partial summary information with already generated partial summary information, and in the second summary information, the partial summary information of the second processing section and the partial summary information of the first processing section may be integrated.
[0013] The information processing system of the present invention includes an editing request in which the prompt integrates the newly generated partial summary information and the already generated partial summary information when the content of the newly generated partial summary information and the content of the already generated partial summary information are related, and in the second summary information, the partial summary information of the second processing section and the partial summary information of the first processing section may be integrated when the content of the partial summary information generated for the second processing section is related to the content of the partial summary information of the first processing section.
[0014] The information processing system of the present invention includes a character limit on the prompt for the summary information, and the second summary information may be generated in accordance with the character limit.
[0015] The information processing system of the present invention further comprises a presentation unit, wherein the prompt includes a question generation request that generates a question request to the speaker to add missing information when the generation model lacks the information necessary to generate the summary information, the dialogue unit acquires the question when the question is generated, and the presentation unit may present the question to the user.
[0016] The information processing system of the present invention includes a prompt that includes a request to add a timestamp indicating the timing of an utterance corresponding to the summary information to the summary information, and the partial summary information may have a timestamp added to it.
[0017] The present invention is a program for causing a computer to function as an acquisition unit and an interaction unit, wherein the acquisition unit acquires user utterance information, and the interaction unit inputs the utterance information to a generation model along with a prompt that includes a summary generation request to generate summary information including partial summary information of the utterance information of a predetermined processing section whenever utterance information of the processing section is obtained, and an editing request to edit already generated summary information based on newly obtained partial summary information of the processing section, thereby acquiring first summary information including partial summary information of a first processing section, and after acquiring the first summary information, acquiring second summary information including partial summary information of the first processing section and partial summary information of a second processing section after the first processing section, and the partial summary information of the first processing section included in the second summary information is edited based on the partial summary information of the second processing section.
[0018] The present invention is an information processing method performed by a computer having a control unit, comprising the steps of: the control unit acquiring user speech information; the control unit inputting the speech information to a generation model along with a prompt that includes a summary generation request to generate summary information including partial summary information of the speech information of a predetermined processing section each time speech information of a predetermined processing section is obtained, and an editing request to edit already generated summary information based on the newly obtained partial summary information of the processing section, thereby acquiring first summary information including partial summary information of a first processing section; and after acquiring the first summary information, acquiring second summary information including partial summary information of the first processing section and partial summary information of a second processing section after the first processing section, wherein the partial summary information of the first processing section included in the second summary information is edited based on the partial summary information of the second processing section.
[0019] According to the information processing system, program, and information processing method of the present invention, it is possible to present summarized information that is easy to understand regarding the content of spoken utterances.
[0020] This is an overall diagram of the summary generation system. This is a flowchart showing the processing of the summary generation device. This is a diagram showing an example of summary information. This is a diagram showing an example of summary information. This is a diagram showing an example of summary information.
[0021] Embodiments of the present invention will be described below with reference to the drawings.
[0022] Figure 1 is an overall diagram of the information processing system 1 according to this embodiment. The information processing system 1 generates summary information for input speech information using a generation model. For example, the information processing system 1 generates a summary of the content from the speech information of a speaker at a lecture or the speech information of an instructor at a lecture, or generates minutes from the speech information of participants at a meeting. The speech information may be in the form of audio data or text data.
[0023] The information processing system 1 comprises an information processing device 10, a generation model server device 20, and a user device 30. The information processing device 10 is an example of an information processing system. The information processing device 10, the generation model server device 20, and the user device 30 are connected to each other so as to be able to communicate via a communication network 2 such as the Internet.
[0024] The information processing device 10 is a computer or the like and uses interaction with a generation model 210 stored in a generation model server device 20 to obtain summary information from speech information and present it. The information processing device 10 is composed of a computer or the like and includes a control unit 100, a storage unit 110, and a communication unit 120.
[0025] The control unit 100 includes a processor such as a CPU (Central Processing Unit) and controls the operation of the information processing device 10. The storage unit 110 includes, for example, an HDD (Hard Disk Drive), RAM (Random Access Memory), ROM (Read Only Memory), and SSD (Solid State Drive). Furthermore, the storage unit 110 is not limited to one built into the information processing device 10, but may be a storage medium that can be detachably attached to the information processing device 10 (for example, a USB memory). In this embodiment, the storage unit 110 stores various programs executed by the control unit 100 and various data.
[0026] The communication unit 120 includes a communication interface for communicating with external devices wirelessly or via wired connections. The control unit 100 transmits and receives data with the generation model server device 20 and the user device 30 via the communication unit 120.
[0027] The generative model server device 20 is a computer or the like. The generative model server device 20 stores a generative model 210 that takes a question (prompt) as input and outputs an answer. The generative model 210 is a learning model generated by machine learning and generates an answer corresponding to the prompt. The generative model can be a Generative Pre-trained Transformer or a language model based thereon. Examples of generative models 210 include ChatGPT, GPT-3, GPT-4 by OpenAI, and GPT-J by EleutherAI.
[0028] The generative model server device 20 is used for large-scale natural language processing tasks, such as automatically performing machine translation, text generation, document classification, and question answering. The generative model 210 learns from a vast amount of data and then makes predictions for new inputs. The generative model server device 20 hosts the generative model 210, calls the generative model 210 in response to requests, and returns the results. In this embodiment, the generative model server device 20 receives a prompt from the information processing device 10 and generates a response sentence by performing calculations. The generated response sentence is then transmitted to the information processing device 10.
[0029] The user device 30 is an information processing device. The user device 30 may also be a portable device such as a smartphone, PC, or tablet. The user device 30 comprises a control unit 310, a storage unit 320, a communication unit 330, an operation unit 340, and a display unit 350. The control unit 310, storage unit 320, and communication unit 330 are the same as the control unit 100, storage unit 110, and communication unit 120 of the information processing device 10, respectively. The communication unit 330 includes a communication interface for communicating with external devices wirelessly or via wired connection. The operation unit 340 accepts user input. Examples of the operation unit 340 include a keyboard and a mouse. The display unit 350 is, for example, a monitor and displays various screens. The operation unit 340 and the display unit 350 may be integrated as a touch panel.
[0030] In this embodiment, the user device 30 is assumed to have an application for summary generation installed. This application is used to send and receive information with the information processing device 10.
[0031] Next, the configuration of the control unit 100 of the information processing device 10 will be described. The control unit 100 functions as an acquisition unit 101, an interaction unit 102, and a presentation unit 103 by executing a program stored in the storage unit 110. In the following, the processes described as being performed by the acquisition unit 101, the interaction unit 102, and the presentation unit 103 are processes performed by the control unit 100 by executing a program.
[0032] The acquisition unit 101 acquires the speech information to be used for summary generation from the user device 30 via the communication unit 120.
[0033] The dialogue unit 102 obtains the response text obtained by the generation model 210 from the generation model server device 20 by sending a prompt to the generation model server device 20 via the communication unit 120. The dialogue unit 102 sends, for example, a prompt requesting the generation of a summary of the utterance information along with the utterance information. As a result, the dialogue unit 102 obtains summary information corresponding to the utterance information.
[0034] Specifically, the dialogue unit 102 inputs speech information generated in real time, for example, during a meeting, into the sequential generation model 210. The dialogue unit 102 then causes the generation model 210 to sequentially generate summary information for the speech information obtained in predetermined processing intervals. Here, the processing interval is a fixed period, such as two minutes. The processing interval is variable and can be set by the user. Hereinafter, the summary information corresponding to a processing interval will be referred to as partial summary information. When partial summary information for multiple processing intervals is obtained, the generation model 210 generates summary information by combining these. Here, the summary information is summary information for the entirety of the processing intervals obtained so far (or multiple processing intervals if multiple processing intervals have been obtained).
[0035] For example, four minutes after the start of the meeting, partial summary information corresponding to each of the two processing intervals is generated, and based on these, a single summary is generated. Then, six minutes after the start of the meeting, further partial summary information corresponding to a third processing interval is generated, and based on the summary information already generated for the two processing intervals and the newly generated partial summary information, a summary of the first six minutes of the meeting is generated. In other words, the summary information is updated every two minutes.
[0036] The presentation unit 103 presents the summary information obtained by the generation model 210 to the user. Specifically, the presentation unit 103 transmits the summary information to the user device 30 via the communication unit 120 and displays it on the display unit 350. In another example, the user device 30 may be equipped with a speaker, and the presentation unit 103 may output the audio data of the summary information from the speaker.
[0037] Figure 2 is a flowchart illustrating the processing performed by the information processing device 10. Here, we will explain using the example of presenting meeting minutes (summary information) for a meeting conducted by multiple people. Before the start of the meeting, the dialogue unit 102 sends a prompt to the generation model server device 20 (step S100). The prompt sent to the generation model server device 20 is input to the generation model 210. Here, we will explain the prompt sent by the dialogue unit 102.
[0038] The prompt includes a summary generation request. This summary generation request includes a request to generate a partial summary of the utterance information for a processing interval whenever utterance information for that interval is obtained. The summary generation request also includes a request to generate new summary information based on the previously obtained summary information and the newly obtained partial summary information, if new partial summary information for a processing interval is obtained. Furthermore, the summary generation request includes information to limit the number of characters in the summary information.
[0039] The prompt also includes edit requests. Edit requests are requests to edit already generated summary information. Edit requests include requests to modify the generated summary information based on newly generated partial summary information. Edit requests also include requests not to add newly generated partial summary information to the summary information if the content of the newly generated partial summary information differs from the content of the generated summary information. Edit requests also include requests to merge newly generated partial summary information with already generated partial summary information. More specifically, requests to merge newly generated partial summary information with already generated partial summary information if the content of the newly generated partial summary information and the content of the already generated partial summary information are related.
[0040] The prompt also includes a request for information. The request for information includes a request to include in the summary information an indication that the partial summary information has been removed, if the partial summary information has been removed.
[0041] The prompt also includes a question generation request. The question generation request includes a request to generate a question that asks the speaker to provide the missing information when the generation model 210 lacks the information necessary to generate summary information.
[0042] Examples of situations where information is insufficient include discrepancies in understanding among multiple speakers, inconsistencies in the content of spoken information, and ambiguous expressions.
[0043] The prompt also includes a granting request. The granting request includes a request to assign a timestamp to the summary information. The timestamp may be information indicating the time when the utterance information was obtained, or information indicating the elapsed time from the start time of recording the utterance information.
[0044] Furthermore, the prompt includes a reference request. The reference request includes a request to refer to the prerequisite information for the utterance information to be input hereafter. The prerequisite information may be relevant summary information or information previously input by the user. Examples of relevant summary information include summary information generated in past meetings held for the same agenda as the target meeting. The prerequisite information includes the agenda of the target meeting, the number of participants, attributes, and the like.
[0045] After transmitting the prompt or together with the prompt, the dialogue unit 102 transmits the prerequisite information to the generation model server device 20 (step S102). The prompt and the prerequisite information are input into the generation model 210. Next, when the meeting starts and the recording of the utterance information starts, the dialogue unit 102 sequentially obtains the utterance information from the user device 30 in real time and sequentially transmits this to the generation model server device 20 (step S104).
[0046] The generation model 210 sequentially receives the utterance information, generates partial summary information in processing section units according to the prompt, and generates summary information based on the partial summary information. The summary information is transmitted from the generation model server device 20 to the information processing device 10.
[0047] When the dialogue unit 102 receives the summary information (Y in step S106), the presentation unit 103 presents the summary information to the user (step S108). Specifically, the presentation unit 103 causes the summary information to be displayed on the display unit 350 of the user device 30. Next, the control unit 100 checks whether the acquisition of the utterance information has ended (step S110). If the acquisition of the utterance information has not ended (N in step S110), the control unit 100 advances the process to step S106. If the acquisition of the utterance information has ended (Y in step S110), the control unit 100 ends the process.
[0048] FIG. 3A, FIG. 3B, and FIG. 3C are all diagrams showing an example of summary information. FIG. 3A is a diagram showing an example of summary information obtained in the first processing section (the first processing section). Here, it is assumed that the processing section is 2 minutes.
[0049] The summary information shown in FIG. 3A is summary information generated from the utterance information in the first processing section. The summary information in the first processing section is equal to the partial summary information in the first processing section. The summary information in the first processing section includes four items of (item 1-1) to (item 1-4) of the "main items", and detailed content for each item is also included. These information are the partial summary information in the first processing section generated in response to the summary generation request included in the prompt. In addition, in the generation of the partial summary information in the first processing section, the prerequisite information is referred to in response to the reference request. Depending on the fact that the number of participants, attributes, and names of the participants are clarified in the prerequisite information, as shown in FIG. 3A, the names of the participants, etc. are described in the summary information in the first processing section. Thus, in the first processing section, although the amount of information in the utterance information is small, since the prerequisite information is referred to, summary information that supplements the content of the utterance information is generated.
[0050] Furthermore, in response to the request for attaching a timestamp, a timestamp is attached to each main item. For example, for (item 1-1), a timestamp of "40 seconds after the start" is attached. Thus, since a timestamp is attached, the user can confirm the corresponding utterance information relying on the timestamp.
[0051] The summary information shown in FIG. 3A includes a question sentence for the participants (speakers) of the meeting in the "question" item in response to the question sentence generation request. For example, (question 1-1) is a question sentence for clarifying an ambiguous expression. Also, (question 1-2) is a question sentence for clarifying unclear content.
[0052] Thus, if the utterance information contains ambiguous or unclear content, the summary information includes a question to clarify this content. The question included in the summary information only needs to request the speaker to add the missing information when the generation model 210 lacks sufficient information to generate partial summary information, and its specific content is not limited to the embodiment.
[0053] Another example is when speaker A says that the synthesis of compound X failed, and speaker B says that the synthesis of compound X was successful. In this case, a question such as "Were you successful in synthesizing compound X?" is generated. In this way, when there is a discrepancy or contradiction in the content of the spoken information, a question may be generated to resolve the discrepancy or contradiction.
[0054] Furthermore, the presentation of the question is not limited to display on the display unit 350. As another example, if the user device 30 is equipped with a speaker, the question may be output as audio from the speaker.
[0055] Figure 3B shows an example of summary information corresponding to the second processing section following the first processing section. In Figure 3B, the parts that have been changed or added from the summary information of the first processing section shown in Figure 3A are underlined. The summary information for the second processing section is generated in response to a summary generation request, based on the summary information of the first processing section and the partial summary information of the second processing section.
[0056] In the "Summary" section, the following has been added: "It was shared that a good yield (85%) was achieved. However, Mr. Sato asked a question regarding the separation and purification status, and Mr. Yamada withheld his answer." Furthermore, in the "Main Items" section, "(Item 2-1) Separation and Purification Status" has been newly added to the summary information. These correspond to the partial summary information for the second processing interval.
[0057] For example, if, each time speech information for a new processing section is obtained, new summary information is generated from the speech information of the processing sections obtained so far, including the new processing section, then significant discrepancies in content may occur between the summary information obtained each time a processing section is added. In contrast, in the information processing system 1 of this embodiment, new summary information is generated based on the summary information that has already been generated and the newly generated partial summary information. This prevents significant discrepancies in content between the summary information created up to the previous time and the newly generated summary information.
[0058] Furthermore, if an answer to the question presented in the summary information of the first processing section is obtained during the meeting in the second processing section, it will be marked as "Resolved," as shown in (Question 1-1) under "Question" in Figure 3B. In addition, the resolved content will be reflected in "Summary," "Main Items," and "Details."
[0059] In this way, the summary information is updated every two minutes, allowing meeting participants to check the content of the questions as needed during the meeting. Therefore, the meeting can proceed while clarifying any points of confusion.
[0060] Figure 3C shows an example of summary information corresponding to the ninth processing interval. In Figure 3C, the parts that have been changed or added from the summary information for the second processing interval shown in Figure 3B are underlined. Also, the parts that have been deleted from the summary information for the second processing interval are struck through.
[0061] In the summary information for the ninth processing interval, items 1-3 of the "main items" included in the summary generation request are deleted, and items 4-1, 5-1, and 6-1 are newly added. The longer the utterance information, the more information is included in the summary. However, the amount of information is limited according to the character limit entered as a prompt. Accordingly, partial summary information included in the summary information may be deleted as appropriate.
[0062] Furthermore, item 6-1 is assigned timestamps of "9 minutes and 19 seconds after the start" and "17 minutes and 34 seconds after the start." This indicates that the partial summary information generated in response to the utterance information at "17 minutes and 34 seconds after the start" has been integrated into item 6-1, "Technical study of reaction temperature control," which is the partial summary information generated in response to the utterance information at "9 minutes and 19 seconds after the start." In this case, for example, the "Details" column may contain content corresponding to the utterance information at "9 minutes and 19 seconds after the start" and content corresponding to the utterance information at "17 minutes and 34 seconds after the start."
[0063] In this way, unnecessary information is deleted or merged in the summary information. Therefore, it is possible to prevent the summary information from becoming too long. The merging of two partial summary information may be performed only when the content of both partial summary information is related. This makes the summary information easier to read.
[0064] Partial summary information may be edited upon request. For example, partial summary information obtained in a certain processing interval P may be changed according to partial summary information obtained in a processing interval Q that follows P. More specifically, suppose partial summary information of "80% yield" is obtained in a certain processing interval P. If, in a subsequent processing interval Q, the utterance information "The yield was 70%, not 80%" is obtained, then the partial summary information in processing interval P is changed from "80% yield" to "70% yield".
[0065] Furthermore, suppose that in a certain processing interval P, partial summary information is obtained stating "Yield data could not be obtained." If, in the subsequent processing interval R, utterance information is obtained stating "It was incorrect to say that yield data could not be obtained; the yield was 80%," then the partial summary information "Yield data could not be obtained" is deleted.
[0066] The prompt's editing request may include a request not to add the newly generated partial summary information to the summary information if the content of the newly generated partial summary information differs from the content of the previously generated summary information. Accordingly, for example, if the content of the partial summary information for a certain processing section T differs from the content of the summary information corresponding to processing sections S prior to processing section T, the partial summary information for processing section T will not be included in the summary information corresponding to processing section T. This makes it possible to exclude, for example, topics or small talk that deviate from the main subject of the meeting from the summary information.
[0067] Furthermore, if partial summary information is deleted from the summary information in response to a prompt, the summary information may include a statement indicating that the partial summary information has been deleted, such as "The content of... has been deleted." This allows the user to understand that the partial summary information has been deleted.
[0068] Furthermore, deleted content may be redisplayed at the request of the user (meeting participant). For example, if a meeting participant says, "Please display the deleted content...", the dialogue unit 102 inputs a prompt to the generation model 210 that includes a request to add partial summary information of the deleted processing section. In this case, the generation model 210 adds the partial summary information of the deleted processing section to the summary information of the processing section following the processing section containing the utterance, "Please display the deleted content...". In this way, the deleted partial summary information can also be presented to the meeting participant at their request. Note that the method of presenting the deleted partial summary information is not limited to this embodiment. As another example, the deleted partial summary information may be displayed on the display unit 350 as information separate from the summary information, such as a list of deleted content.
[0069] Furthermore, the summary information is updated each time a processing interval has elapsed. Therefore, meeting participants can review the questions during the meeting. As a result, if answers to the questions are spoken during the meeting, they can be reflected in the summary information.
[0070] As described above, the information processing system 1 of this embodiment can present summary information that is easy to understand regarding the content of the utterance.
[0071] The embodiments described above are merely examples for carrying out the present invention, and various other embodiments can be adopted. For example, various modifications and changes are possible within the scope of the gist of the present invention as described in the claims, such as applying one modification to another. For example, some of the components of the above embodiments may be omitted, or the order of processing may be changed or omitted.
[0072] As a first variation of this, the dialogue unit 102 may present the user with information regarding time management if the background information or initial utterance information includes the scheduled time required for the utterance information, such as a meeting. For example, if the meeting is scheduled to last one hour, the dialogue unit 102 may present announcements such as, "30 minutes have passed," or "There are 10 minutes left until the scheduled end time." Furthermore, if the agenda is known, the dialogue unit 102 may also present announcements to assist in the progress of the meeting, such as, "30 minutes have passed, but no conclusions have been reached on agenda items V and W."
[0073] A second variation is that the character limit for the summary information may be changed during the meeting. For example, during the meeting, depending on the status of updates to the summary information, there may be a desire to increase or decrease the character count of the summary information. In such cases, the character count in the summary information can be changed during the meeting by prompting the generation model 210 via the dialogue unit 102 to change the character count.
[0074] According to the information processing system, program, and information processing method of this embodiment, which have the above configuration, the information processing system is an information processing system 1 comprising an acquisition unit 101 and a dialogue unit 102, wherein the acquisition unit 101 acquires user utterance information, and the dialogue unit 102 inputs utterance information to a generation model 210 along with a prompt that includes a summary generation request to generate summary information including partial summary information of the utterance information of a predetermined processing section whenever utterance information of a predetermined processing section is obtained, and an editing request to edit the already generated summary information based on the newly obtained partial summary information of the processing section, thereby acquiring first summary information including partial summary information of a first processing section, and after acquiring the first summary information, acquiring second summary information including partial summary information of the first processing section and partial summary information of a second processing section after the first processing section, and the partial summary information of the first processing section included in the second summary information is edited based on the partial summary information of the second processing section. In this way, in the second summary information, the partial summary information of the first processing section is edited, making it easier to understand the utterance content.
[0075] Furthermore, according to the information processing system, program, and information processing method of this embodiment, the prompt includes an editing request to modify the generated summary information based on newly generated partial summary information, and the second summary information includes the partial summary information of the first processing section modified based on the partial summary information of the second processing section. In this way, since the partial summary information of the first processing section has been edited in the second summary information, the content of the utterance is easier to understand.
[0076] Furthermore, according to the information processing system, program, and information processing method of this embodiment, the prompt includes an editing request to delete the previously generated partial summary information based on the newly generated partial summary information, and the second summary information does not include the partial summary information of the first processing section. In this way, since the second summary information does not include the partial summary information of the first processing section, it is possible to prevent the summary information from becoming redundant.
[0077] Furthermore, according to the information processing system, program, and information processing method of this embodiment, the prompt includes a request to include information indicating that partial summary information has been deleted in the summary information when partial summary information is deleted, and the second summary information includes information indicating that partial summary information for the first processing section has been deleted. This allows the user to understand that partial summary information has been deleted.
[0078] Furthermore, according to the information processing system, program, and information processing method of this embodiment, if the user requests display of deleted partial summary information in the second summary information, the dialogue unit inputs a prompt to the generation model that includes a request to add the deleted partial summary information, thereby obtaining the deleted partial summary information of the first processing section. This prevents the summary information from becoming redundant.
[0079] Furthermore, according to the information processing system, program, and information processing method of this embodiment, the prompt includes an editing request that, if the content of the newly generated partial summary information differs from the content of the already generated summary information, the newly generated partial summary information is not added to the summary information. If the content of the partial summary information for a third processing section that follows the second processing section differs from the content of the second summary information, the third summary information generated after the second summary information does not include the partial summary information for the third processing section. This makes it possible to avoid including unnecessary information in the summary information.
[0080] Furthermore, according to the information processing system, program, and information processing method of this embodiment, the prompt includes an editing request that integrates newly generated partial summary information with already generated partial summary information, and in the second minutes information, the partial summary information of the second processing section and the partial summary information of the first processing section are integrated. This makes it possible to generate summary information that is easier to understand.
[0081] Furthermore, according to the information processing system, program, and information processing method of this embodiment, a prompt includes an editing request to integrate newly generated partial summary information and already generated partial summary information when the content of newly generated partial summary information and the content of already generated partial summary information are related, and when the content of partial summary information generated for a second processing section is related to the content of partial summary information for a first processing section, the partial summary information for the second processing section and the partial summary information for the first processing section are integrated in the second summary information. This makes it possible to generate summary information that is easier to understand.
[0082] Furthermore, according to the information processing system, program, and information processing method of this embodiment, the prompt includes a character limit for the summary information, and the second summary information is generated according to the character limit. Therefore, it is possible to prevent the summary information from becoming too long.
[0083] Furthermore, according to the information processing system, program, and information processing method of this embodiment, a presentation unit is further provided, and the prompt includes a question generation request that the generation model generate a question request that the speaker add the missing information when there is insufficient information to generate summary information, the dialogue unit retrieves the question when a question is generated, and the presentation unit presents the question to the user. This makes it possible to obtain better summary information.
[0084] Furthermore, according to the information processing system, program, and information processing method of this embodiment, the prompt includes an assignment request that assigns a timestamp indicating the timing of the utterance corresponding to the summary information to the summary information, and the partial summary information is assigned a timestamp. This allows the user to understand which part of the utterance information the partial summary information corresponds to.
[0085] 1 Information Processing System 10 Information Processing Device 20 Generation Model Server Device 30 User Device 100 Control Unit 101 Acquisition Unit 102 Dialogue Unit 103 Presentation Unit 110 Storage Unit 120 Communication Unit 210 Generation Model 310 Control Unit 320 Storage Unit 330 Communication Unit 340 Operation Unit 350 Display Unit
Claims
1. An information processing system comprising an acquisition unit and a dialogue unit, wherein the acquisition unit acquires user utterance information, and the dialogue unit inputs the utterance information to a generation model along with a prompt that includes a summary generation request to generate summary information including partial summary information of the utterance information of a predetermined processing section whenever utterance information of a predetermined processing section is obtained, and an editing request to edit already generated summary information based on newly obtained partial summary information of a processing section, thereby acquiring first summary information including partial summary information of a first processing section, and after acquiring the first summary information, acquiring second summary information including partial summary information of the first processing section and partial summary information of a second processing section after the first processing section, and the partial summary information of the first processing section included in the second summary information is edited based on the partial summary information of the second processing section.
2. The information processing system according to claim 1, wherein the prompt includes an editing request to modify generated summary information based on newly generated partial summary information, and the second summary information includes partial summary information of the first processing section modified based on the partial summary information of the second processing section.
3. The information processing system according to claim 1, wherein the prompt includes an editing request to delete generated partial summary information based on newly generated partial summary information, and the second summary information does not include partial summary information of the first processing interval.
4. The information processing system according to claim 3, wherein the prompt includes a request to include in the summary information information indicating that the partial summary information has been deleted when the partial summary information is deleted, and the second summary information includes information indicating that the partial summary information of the first processing interval has been deleted.
5. In the second summary information, if the user requests the display of the deleted partial summary information, the dialogue unit inputs a prompt to the generation model including a request to add the deleted partial summary information, thereby obtaining the deleted partial summary information of the first processing section, as described in claim 3.
6. The information processing system according to any one of claims 1 to 5, wherein the prompt includes an editing request not to add the newly generated partial summary information to the summary information if the content of the newly generated partial summary information differs from the content of the previously generated summary information, and the third summary information generated after the second summary information does not include the partial summary information of the third processing section if the content of the partial summary information of a third processing section after the second processing section differs from the content of the second summary information.
7. The information processing system according to any one of claims 1 to 6, wherein the prompt includes the editing request for integrating newly generated partial summary information with already generated partial summary information, and in the second summary information, the partial summary information of the second processing section and the partial summary information of the first processing section are integrated.
8. The information processing system according to claim 7, wherein the prompt includes an editing request to integrate the newly generated partial summary information and the already generated partial summary information when the content of the newly generated partial summary information and the content of the already generated partial summary information are related, and the second summary information integrates the partial summary information of the second processing section and the partial summary information of the first processing section when the content of the partial summary information generated for the second processing section is related to the content of the partial summary information of the first processing section.
9. The information processing system according to any one of claims 1 to 8, wherein the prompt includes a character limit for the summary information, and the second summary information is generated in accordance with the character limit.
10. An information processing system according to any one of claims 1 to 9, further comprising a presentation unit, wherein the prompt includes a question generation request that generates a question request that the speaker provide the missing information when the generation model lacks the information necessary to generate the summary information, the dialogue unit acquires the question when the question is generated, and the presentation unit presents the question to the user.
11. The information processing system according to any one of claims 1 to 10, wherein the prompt includes a request to assign a timestamp indicating the timing of an utterance corresponding to the summary information to the summary information, and the partial summary information has a timestamp assigned to it.
12. A program for causing a computer to function as an acquisition unit and an interaction unit, wherein the acquisition unit acquires user utterance information, and the interaction unit inputs the utterance information to a generation model along with a prompt that includes a summary generation request to generate summary information including partial summary information of the utterance information of a predetermined processing section whenever utterance information of a predetermined processing section is obtained, and an editing request to edit already generated summary information based on newly obtained partial summary information of the processing section, thereby acquiring first summary information including partial summary information of a first processing section, and after acquiring the first summary information, acquiring second summary information including partial summary information of the first processing section and partial summary information of a second processing section after the first processing section, and the partial summary information of the first processing section included in the second summary information is edited based on the partial summary information of the second processing section.
13. An information processing method performed by a computer having a control unit, comprising: a step of the control unit acquiring user speech information; a step of the control unit inputting the speech information to a generation model, each time speech information for a predetermined processing section is obtained, along with a prompt including a summary generation request that generates summary information including partial summary information of the speech information for the processing section, and an editing request that edits already generated summary information based on the newly obtained partial summary information of the processing section, thereby acquiring first summary information including partial summary information of a first processing section; and after acquiring the first summary information, acquiring second summary information including partial summary information of the first processing section and partial summary information of a second processing section after the first processing section, wherein the partial summary information of the first processing section included in the second summary information is edited based on the partial summary information of the second processing section.