Operating method, information processing device

Abstract

Improving the accuracy of summarizing dialogue content based on audio data of conversations. [Solution] Based on the audio data of the conversation between the salesperson and the customer, the information processing device 20 presents a summary of the past topic, the third topic, whenever the current topic of the conversation, the first topic, changes.

Description

Technical Field

[0001] This disclosure relates to an operation method and an information processing apparatus.

Background Art

[0002] Techniques for summarizing conversation content from conversation voice data are known. For example, Patent Document 1 discloses a conversation analysis system that records conversation data based on voice data of a recorded conversation, extracts conversation content that matches conditions specified by a user from the conversation data, and displays it in a list.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] With the development of machine learning and the like, there is room for improvement in the accuracy of summarizing conversation content from conversation voice data.

[0005] In view of such circumstances, an object of the present disclosure is to improve the accuracy of summarizing conversation content based on conversation voice data.

Means for Solving the Problems

[0006] An operation method according to an embodiment of the present disclosure is [[ID=৪৭]] an operation method of an information processing apparatus, acquiring voice data of a conversation between a salesperson and a customer, converting the voice data into text data using voice recognition technology, deriving a first topic, which is the current topic of the conversation, from the text data using co-occurrence analysis or a topic model, Each time a change in the first topic is detected, Identifying the audio data of the dialogue portion related to the immediately preceding topic, the second topic. Based on the identified audio data, generate a summary of the dialogue portion, and To present the generated summary to the salesperson, Includes.

[0007] The operating method according to one embodiment of this disclosure is: A method for operating an information processing device, Based on audio data of conversations between salespeople and customers, whenever the current topic of the conversation (topic 1) changes, a summary of the past topic (topic 3) is presented. Includes.

[0008] An information processing apparatus according to one embodiment of this disclosure is We acquire audio data of conversations between sales staff and customers. Using speech recognition technology, the aforementioned speech data is converted into text data. Using co-occurrence analysis or topic modeling, the first topic, which is the current topic of the dialogue, is derived from the text data. Each time a change in the first topic is detected, Identify the audio data of the dialogue portion related to the immediately preceding topic, the second topic. Based on the identified audio data, a summary of the dialogue portion is generated. The generated summary is presented to the salesperson. It is equipped with a control unit.

[0009] An information processing apparatus according to one embodiment of this disclosure is The system includes a control unit that, based on audio data of a conversation between a salesperson and a customer, presents a summary of a past topic (a third topic) whenever the current topic (a first topic) of the conversation changes. [Effects of the Invention]

[0010] According to an embodiment of the present disclosure, it is possible to improve the accuracy of the summary of the dialogue content based on the voice data of the dialogue.

Brief Description of the Drawings

[0011] [Figure 1] It is a block diagram showing a schematic configuration of a system according to an embodiment of the present disclosure. [Figure 2] It is a block diagram showing a schematic configuration of a terminal device. [Figure 3] It is a block diagram showing a schematic configuration of an information processing device. [Figure 4] It is a flowchart showing the operation of the information processing device.

Modes for Carrying Out the Invention

[0012] Hereinafter, embodiments of the present disclosure will be described.

[0013] (Overview of the Embodiment) Referring to FIG. 1, the overview of the information processing system 1 according to the embodiment of the present disclosure will be described. The information processing system 1 includes one or more terminal devices 10 and an information processing device 20, respectively. The terminal device 10 and the information processing device 20 are communicably connected to a network 30 including, for example, the Internet and a mobile communication network.

[0014] The terminal device 10 is a computer such as a PC (Personal Computer), a smartphone, or a tablet terminal. The terminal device 10 is a computer used by a salesperson in a store such as a vehicle dealership when having a conversation with a customer.

[0015] The information processing device 20 is, for example, one or a plurality of server computers communicable with each other. The information processing device 20 can communicate information with each terminal device 10 via the network 30. The information processing device 20 provides services used by salespersons in stores such as vehicle dealerships.

[0016] In this embodiment, the information processing device 20 presents a summary of the past topic, the third topic (hereinafter referred to as the past topic), each time the first topic (hereinafter referred to as the current topic), which is the current topic of the conversation, changes, based on the audio data of the conversation between the salesperson and the customer. In this embodiment, a topic is a label assigned to words that frequently appear in the conversation. The labels assigned to words may be selected from a set of pre-set candidates. For example, candidate labels for a conversation between a salesperson and a customer regarding a vehicle sale may include "vehicle description," "insurance," "service options for the vehicle," "payment methods," and "legal explanation." A group of words indicating the name of a vehicle, such as "Crown" or "Lexus," may be labeled "vehicle description," and a group of words such as "lump-sum purchase" or "installment purchase" may be labeled "payment methods." The pre-defined label candidates may also include an "other" label for words that do not correspond to any of the candidates.

[0017] When summarizing dialogue content from audio data, problems can arise if the recording data is divided into predetermined time intervals for summarization, or if the transcribed text of the recording data is divided into predetermined character counts for summarization. For example, in relatively long dialogues such as business negotiations, it can be difficult to obtain a highly accurate summary of the entire dialogue by dividing the dialogue into short segments for summarization. In contrast, according to this embodiment, a summary of past topics is presented each time the current topic of the dialogue changes. Therefore, even in relatively long dialogues such as business negotiations, summaries are generated on a topic-by-topic basis during the negotiation. Thus, compared to generating summaries based on time intervals of the recording data or on the amount of text, the accuracy of the summaries can be improved.

[0018] Next, we will describe the various components of Information Processing System 1.

[0019] (Terminal device configuration) As shown in Figure 2, the terminal device 10 includes a communication unit 11, an output unit 12, an input unit 13, a storage unit 14, and a control unit 15.

[0020] The communication unit 11 includes one or more communication interfaces connected to the network 30. The communication interfaces are compatible with, but are not limited to, mobile communication standards such as 4G (4th Generation) or 5G (5th Generation). The terminal device 10 communicates with the information processing device 20 via the communication unit 11 and the network 30.

[0021] The output unit 12 includes one or more output devices for outputting information. These output devices may include, for example, a display for outputting images and a speaker for outputting sound. Alternatively, the output unit 12 may include an interface for connecting external output devices.

[0022] The input unit 13 includes one or more input devices for detecting user input operations. These input devices include, for example, physical keys, capacitive keys, a mouse, a touch panel, a touchscreen integrated with the display of the output unit 12, a microphone, etc. Alternatively, the input unit 13 may include an interface for connecting external input devices.

[0023] The storage unit 14 includes one or more memories. The memories are, for example, semiconductor memories, magnetic memories, or optical memories, but are not limited to these. Each memory included in the storage unit 14 may function as, for example, a main memory, an auxiliary memory, or a cache memory. The storage unit 14 stores any information used for the operation of the terminal device 10. In addition, for example, the storage unit 14 may store system programs, application programs, and embedded software. For example, the information stored in the storage unit 14 may be updatable with information obtained from the network 30 via the communication unit 11. The storage unit 14 may store a list of pre-set label candidates.

[0024] The control unit 15 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or a combination thereof. The processor is a general-purpose processor such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit), or a dedicated processor specialized for a specific process, but is not limited to these. The programmable circuit is an FPGA (Field-Programmable Gate Array), but is not limited to this. The dedicated circuit is an ASIC (Application Specific Integrated Circuit), but is not limited to this. The control unit 15 controls the operation of the terminal device 10.

[0025] (Configuration of information processing device) As shown in Figure 3, the information processing device 20 comprises a communication unit 21, a storage unit 22, and a control unit 23.

[0026] The communication unit 21 includes one or more communication interfaces connected to the network 30. The communication interfaces may support, for example, mobile communication standards, wired LAN (Local Area Network) standards, or wireless LAN standards, but are not limited to these, and may support any communication standard. The information processing device 20 communicates with the terminal device 10 via the communication unit 21 and the network 30, respectively.

[0027] The storage unit 22 includes one or more memories. Each memory included in the storage unit 22 may function as, for example, a main memory, an auxiliary memory, or a cache memory. The storage unit 22 stores any information used in the operation of the information processing device 20. For example, the storage unit 22 may store system programs, application programs, embedded software, and map information. The storage unit 22 may also store a list of pre-set label candidates.

[0028] The control unit 23 includes one or more processors, one or more programmable circuits, one or more dedicated circuits, or a combination thereof. The processor is a general-purpose processor such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit), or a dedicated processor specialized for a specific process, but is not limited to these. The programmable circuit is an FPGA (Field-Programmable Gate Array), but is not limited to this. The dedicated circuit is an ASIC (Application Specific Integrated Circuit), but is not limited to this. The control unit 23 controls the operation of the entire information processing device 20.

[0029] (Operation flow of the information processing device) Referring to Figure 4, the operation of the information processing device 20 according to this embodiment will be described. Each step in Figure 4 is an information processing step executed by the control unit 23 of the information processing device 20. The procedure in Figure 4 is executed by the control unit 23 at arbitrary intervals, such as every 10 milliseconds. The interval at which each step is executed may vary depending on the performance of each component of the information processing system 1, such as the performance of the information processing device 20.

[0030] S100: The control unit 23 of the information processing device 20 acquires voice data of the conversation between the salesperson and the customer. Specifically, the control unit 23 receives voice data from the terminal device 10 via the network 30 using the communication unit 21. The terminal device 10 acquires voice data when the salesperson is negotiating with the customer using the microphone of the input unit 13 or a microphone connected via the input unit 13. Alternatively, the control unit 23 may receive voice data directly from the microphone via the network 30 using the communication unit 21. The voice data is acquired at arbitrary intervals, for example, every 10 milliseconds. The interval at which the voice data is acquired may vary depending on the performance of each device constituting the information processing system 1, such as the performance of the information processing device 20.

[0031] S101: The control unit 23 derives the current topic of the dialogue from the audio data. Specifically, the control unit 23 converts the audio data into text data using speech recognition technology. The speech recognition technology includes any technology such as natural language processing (NLP), hidden Markov models (HMMs), and deep neural networks (DNNs). The control unit 23 then identifies groups of words that frequently appear in close proximity in time in the text data through co-occurrence analysis and determines the current topic from one of the labels pre-assigned to each word. For example, if vehicle names such as "Crown" and "Lexus" frequently appear in the text data, the control unit 23 determines "Vehicle Description," which is pre-labeled for "Crown" and "Lexus," as the current topic. Also, if the words "lump-sum purchase" and "installment purchase" frequently appear in the text data, the control unit 23 determines "Payment Method," which is pre-labeled for "lump-sum purchase" and "installment purchase," as the current topic. If multiple words with different labels appear frequently, the control unit 23 may, for example, sum the word occurrence frequencies, such as the number of occurrences, for each label and determine the label of the word with the highest occurrence frequency as the current topic. Alternatively, the control unit 23 may use a topic model to determine the current topic corresponding to frequently occurring words in the text data. As a topic model, methods such as Supervised Latent Dirichlet Allocation (sLDA) can be employed. Alternatively, the control unit 23 may have a Large Language Model (LLM) identify frequently occurring words in the text data and select the current topic corresponding to the frequently occurring words from the label candidates. The control unit 23 may use an LLM provided by any provider on the cloud, for example, or an LLM stored in the storage unit 22 of the information processing device 20.

[0032] S102: The control unit 23 detects a change in the current topic. Specifically, the control unit 23 compares the current topic extracted in S101 in the current processing cycle with the current topic extracted in S101 in a past processing cycle, for example, the previous processing cycle. If the comparison reveals that the current topics are different (S102-Yes), the control unit 23 proceeds to S103. If it determines that the current topics are the same (S102-No), the control unit 23 proceeds to S106.

[0033] S103: The control unit 23 identifies audio data corresponding to past topics, for example, the most recent current topic. Specifically, the control unit 23 identifies audio data from the start position of the dialogue, or the position where a change in the previous-to-last current topic was detected, to the position where the most recent change in the current topic was detected, as audio data for past topics.

[0034] S104: The control unit 23 generates a summary of the dialogue corresponding to the past topic based on the identified audio data. Specifically, the control unit 23 extracts the portion of the audio data corresponding to the past topic identified in S103 from the text data generated in S101 of the past processing cycle. Next, the control unit 23 generates a summary using the extracted text data as input data, for example, using LLM. The control unit 23 may generate the summary using extractive summarization, generative summarization, or any combination thereof. The control unit 23 may use LLM provided by any provider on the cloud, for example, or it may use LLM stored in the storage unit 22 of the information processing device 20.

[0035] S105: The control unit 23 presents the generated summary. Specifically, the control unit 23 transmits the summary information generated in S104 to the terminal device 10 via the network 30 using the communication unit 21. The control unit 15 of the terminal device 10 then receives the summary information via the communication unit 11 and presents the summary to the salesperson or customer by displaying a screen containing the summary on the display of the output unit 12.

[0036] The control unit 23 may present a group of summaries relating to past topics, including multiple past topics. Specifically, if the number of times a change in the current topic is detected in S102 is two or more, the control unit 23 transmits not only the summary generated in S104 in the current processing cycle, but also the summary information generated in S104 in previous processing cycles. The control unit 15 of the terminal device 10 then presents the group of summaries relating to past topics to the salesperson by displaying a screen on the output unit 12's display showing the group of summaries received via the communication unit 11. Alternatively, the control unit 15 may store the summary information received from the information processing device 20 in the storage unit 14, and then present the group of summaries relating to past topics to the salesperson by displaying a screen on the output unit 12's display showing the stored summaries. Furthermore, the group of summaries relating to past topics presented to the salesperson may be selectable by the salesperson. Specifically, the control unit 15 of the terminal device 10 may display a screen on the display that allows the summary group stored in the storage unit 14 to be selected in a list format, and when a summary is selected by the salesperson, a screen displaying the selected summary may be displayed on the display.

[0037] S106: The control unit 23 determines whether the conversation between the salesperson and the customer has ended. For example, the control unit 23 may determine that the conversation has ended if it does not receive voice data from the terminal device 10 or a microphone installed in the store for a predetermined period of time. The predetermined period can be arbitrarily set, for example, within the range of 60 to 300 seconds. Alternatively, if the conversation with the customer, such as a business negotiation, may be divided into multiple sessions, the control unit 23 may determine that the conversation has ended by receiving information from the terminal device 10 indicating that the conversation has ended. Specifically, the control unit 15 of the terminal device 10 displays a screen related to the conversation on the display. When the salesperson presses a button related to ending the conversation on the screen, for example, the control unit 15 transmits information to the information processing device 20 indicating that the conversation has ended. When the control unit 23 receives information from the terminal device 10 indicating that the conversation has ended, it determines that the conversation has ended. If it is determined that the conversation has ended, the control unit 23 terminates the processing process. Otherwise, the control unit 23 returns to S100. Note that the method for determining whether the conversation has ended is not limited to the above. For example, the control unit 23 may determine that a conversation has ended if it detects a word, phrase, or other element from the conversation between the salesperson and the customer that suggests the conversation has ended.

[0038] As described above, the information processing device 20 presents a summary of past topics each time the current topic of the conversation changes, based on the audio data of the conversation between the salesperson and the customer.

[0039] With this configuration, a summary of past topics is presented each time the current topic of the dialogue changes. Therefore, even in relatively long dialogues such as business negotiations, a summary is generated for each topic during the negotiation, which can improve the accuracy of the summaries compared to conventional methods such as dividing recorded data into predetermined time intervals and summarizing it. Thus, this embodiment makes it possible to improve the accuracy of summaries compared to conventional methods.

[0040] While this disclosure has been described based on the drawings and embodiments, it should be noted that those skilled in the art may make various modifications and alterations based on this disclosure. Therefore, it should be noted that these modifications and alterations are within the scope of this disclosure. For example, the functions, etc., included in each component or step can be rearranged in a logically consistent manner, and multiple components or steps can be combined into one or divided into two.

[0041] For example, in the above-described embodiment, it is also possible to have an embodiment in which the configuration and operation of the information processing device 20 are distributed among multiple computers that can communicate with each other. Alternatively, for example, it is also possible to have an embodiment in which some or all of the components of the information processing device 20 are provided in the terminal device 10. For example, the terminal device 10 may comprise some or all of the components of the information processing device 20.

[0042] Furthermore, for example, in the embodiment described above, if the current topic falls under one of several important topics arbitrarily set in advance, the control unit 23 may present the salesperson with important topics other than the important topic that falls under the current topic (hereinafter referred to as the relevant topic) (hereinafter referred to as the non-relevant topic). Important topics are candidates from among the label candidates that are considered to have a high need to be addressed, and are set arbitrarily in advance. For example, if the conversation between the salesperson and the customer is a negotiation regarding the sale of a vehicle, "vehicle description," "vehicle insurance," "service options to be added to the vehicle," "payment methods," and "explanation of legal matters" may be set as important topics, and information indicating the order in which they should be addressed is attached to each label designated as an important topic. For example, among the important topics, "vehicle description," "vehicle insurance," and "service options to be added to the vehicle" are ordered in the order of "vehicle description," "service options to be added to the vehicle," and "vehicle insurance." When the control unit 23 extracts the current topic in S101, it determines whether the current topic falls under the category of an important topic. The determination of whether the current topic falls under the category of an important topic may be based on whether the current topic appears in accordance with the order in which it is set as an important topic. For example, the control unit 23 can refer to the history of past topics and determine whether the currently determined topic corresponds to the order of important topics. If the current topic is an important topic, the control unit 23 transmits information about the important topic (hereinafter referred to as the relevant topic) to the terminal device 10. When the control unit 15 of the terminal device 10 receives information about the relevant topic from the information processing device 20, it displays a screen on the display showing a list of non-relevant topics excluding the relevant topic. Alternatively, the control unit 15 can display the important topics in a list format and distinguish between relevant topics that have already been discussed in the conversation and non-relevant topics by, for example, graying out the row for the relevant topic or adding a check mark to the right of the row for the relevant topic.

[0043] Labels for important topics may be set by the salesperson or the administrator managing the service. Important topics may differ for each terminal device 10, or they may differ for groups of terminal devices 10, such as by store. Furthermore, important topics may be changeable.

[0044] Furthermore, in the modified version, the control unit 23 may present to the salesperson important topics from among a plurality of pre-set important topics that did not correspond to past topics, i.e., non-applicable topics. Specifically, the control unit 23 determines whether each past topic or group of past topics is an important topic. If an applicable topic exists, the control unit 23 transmits information about the applicable topic to the terminal device 10, causing the terminal device 10 to display the non-applicable topics, excluding the applicable topic, on its display, distinguishing them from the applicable topic. The control unit 23 may also execute this modified version at any time. For example, if the terminal device 10 is powered off or the service on the terminal device 10 is terminated before it is determined in S106 that the dialogue has ended, the control unit 23 may execute this modified version when it detects that the service on the terminal device 10 has started.

[0045] In further variations, each important topic is associated with a keyword. Keywords are arbitrarily set words that should be included in the dialogue from among the words labeled as important topics. In this case, the labeled words may include information indicating whether or not they are keywords. For example, if the dialogue between a salesperson and a customer is a negotiation about the sale of a vehicle, and the important topic is "payment method," then keywords might include "lump sum purchase," "installment purchase," etc.

[0046] If the current topic corresponds to a pre-set important topic, the control unit 23 may present the salesperson with keywords that have been pre-associated with the important topic corresponding to the current topic but are not included in the dialogue. Specifically, when the control unit 23 extracts the current topic in S101, it determines whether the current topic is an important topic. If the current topic is an important topic, the control unit 23 monitors whether any words set as keywords among the words with the label selected as the current topic can be detected from the text data. If a word set as a keyword is detected, the control unit 23 transmits the detected word to the terminal device 10 as detected keyword information. When the control unit 15 of the terminal device 10 receives the detected keyword information from the information processing device 20, it displays a screen on the display showing a list of words that have been set as keywords excluding the detected keywords (hereinafter referred to as undetected keywords). Alternatively, the control unit 15 may display a screen on the display showing a list of keywords associated with important topics, and may use any method to indicate detected keywords in the keyword section, such as graying out the rows of detected keywords or placing a checkmark at the right end of the rows of detected keywords.

[0047] In a further modification, the control unit 23 may derive and store the emotional states of the salesperson and the customer based on the audio data of the dialogue portion relating to past topics whenever the current topic changes. The emotional state in the dialogue includes emotional states such as joy, anger, sadness, happiness, positive, or negative during the dialogue. Specifically, in S103, when the control unit 23 identifies the audio data of the dialogue portion corresponding to a past topic, for example, the previous current topic, it uses the audio data as input data and derives the emotional states of the salesperson and the customer using a module such as voice emotion recognition stored, for example, on the cloud or in the storage unit 22 of the information processing device 20. The control unit 23 then stores the derived states of the salesperson and the customer in the storage unit 22 or the storage unit 14 of the terminal device 10. At this time, the control unit 23 may also store the summary generated in S104 in the storage unit 22 or storage unit 14. By storing the emotional states of the salesperson and the customer during the dialogue in this way, it can be used, for example, as training data for a learning model using AI. Furthermore, the control unit 23 may use not only the audio data but also the text data and / or image data corresponding to the audio data of the dialogue portion extracted in S104 to detect the emotional states of the salesperson and the customer. When image data is used, the image data may be acquired by an imaging unit mounted on the terminal device 10, or, for example, by a surveillance camera capable of imaging the inside of the store. Performing multimodal analysis in this way can improve the accuracy of deriving the emotional states of the salesperson and the customer.

[0048] Furthermore, in a further modification, the control unit 23 may present the salesperson with the next topic to be discussed based on the summary. Specifically, the control unit 23 uses the summary generated in S104 as input data and, for example, uses LLM to obtain information on the next topic to be discussed. The control unit 23 transmits the topic obtained from LLM to the terminal device 10 via the communication unit 21 and the network 30. When the control unit 15 of the terminal device 10 receives the information on the next topic to be discussed from the information processing device 20, it displays a screen on the display showing the next topic to be discussed. Alternatively, if the label contains information indicating the order in which to discuss the topics, the control unit 23 may determine the next topic to be discussed based on the information indicating the order in which to discuss the topics. For example, suppose the order in which the labels "Vehicle Description" and "Service Options for Vehicles" are discussed is as shown above, and the current topic is "Vehicle Description". In this case, the control unit 23 presents the salesperson with "Service Options for Vehicles" as the next topic to be discussed.

[0049] Furthermore, in a further modification, if the current topic is a pre-set trending topic, the control unit 23 may present the salesperson with a past summary or corresponding manual related to that trending topic. Trending topics are topics that are likely to be brought up by the customer during the conversation between the salesperson and the customer, and are arbitrarily set in advance. In this case, the label includes information indicating whether or not it is a trending topic. In addition, corresponding manuals may be associated with labels set as trending topics. For example, current events such as traffic accident information and gasoline prices may be set as trending topics. Also, for example, matters that have been asked by the customer may be set as trending topics. Specifically, when the control unit 23 extracts the current topic in S101, it determines whether or not it is a trending topic. If the current topic is a trending topic, the control unit 23 transmits information on past summaries or corresponding manuals related to the label that is the current topic to the terminal device 10. When the control unit 15 of the terminal device 10 receives information on past summaries or corresponding manuals from the information processing device 20, it displays the past summaries or corresponding manuals on the display. For summaries concerning trending topics, a summary in which the salesperson and customer states are stored in memory unit 22 or memory unit 14 can be used, as described in the above modifications. Specifically, the control unit 23 searches for past summaries of trending topics from memory unit 22 or memory unit 14. If a summary is found, the control unit 23 determines whether to use the summary based on the emotional states of the salesperson and customer associated with the retrieved summary. For example, if the emotional states of the salesperson and customer include positive reactions, the control unit 23 decides to use the summary as a past summary to present to the salesperson.

[0050] The setting of labels for trending topics may be done by a salesperson or an administrator managing the service. Trending topics may differ for each terminal device 10, or they may differ for groups of terminal devices 10, such as by store. Trending topics may also be changeable.

[0051] Furthermore, in a further modification, the control unit 23 may determine that the current topic has changed if the current topic has not changed for a predetermined amount of time or for a predetermined amount of text (e.g., 3000 characters) or more. Specifically, the control unit 23 starts a timer when a conversation between the salesperson and the customer begins, or when the current topic changes in S102. Then, the control unit 23 determines that the current topic has changed when the timer has exceeded a predetermined amount of time, such as 10 minutes. Alternatively, if the current topic has not changed for a predetermined amount of time, the control unit 23 may send information prompting the termination of the current topic to the terminal device 10. When the control unit 15 of the terminal device 10 receives information prompting the termination of the current topic from the information processing device 20, it displays a screen on the display showing the information prompting the termination of the current topic. The information prompting the termination of the current topic is information presented by the control unit 23 to prompt the salesperson to change the current topic. For example, messages such as "Please change the topic" or "Please move on to the topic of vehicle insurance." By presenting this kind of information to salespeople and having them change the topic of conversation before there is a high probability that the accuracy of the summary will deteriorate, the accuracy of the summary can be maintained.

[0052] Each of the above-described modifications may be performed by the control unit 15 of the terminal device 10, rather than by the control unit 23.

[0053] Furthermore, it is also possible to implement an embodiment in which a general-purpose computer functions as the information processing device 20 according to the above-described embodiment. Specifically, a program describing the processing content that realizes each function of the information processing device 20 according to the above-described embodiment is stored in the memory of the general-purpose computer, and the processor reads and executes the program. Therefore, this disclosure can also be implemented as a program that can be executed by a processor, or as a non-temporary computer-readable medium that stores the program.

[0054] Some embodiments of the present disclosure are described below. However, it should be noted that the embodiments of the present disclosure are not limited to these. [Note 1] A method for operating an information processing device, To acquire audio data of conversations between salespeople and customers, Using speech recognition technology, convert the aforementioned speech data into text data. Using co-occurrence analysis or topic modeling, derive the first topic, which is the current topic of the dialogue, from the text data. Each time a change in the first topic is detected, Identifying the audio data of the dialogue portion related to the immediately preceding topic, the second topic. Based on the identified audio data, generate a summary of the dialogue portion, and To present the generated summary to the salesperson, The method of operation, including the method of operation. [Note 2] A method for operating an information processing device, Based on audio data of conversations between salespeople and customers, whenever the current topic of the conversation (topic 1) changes, a summary of the past topic (topic 3) is presented. The method of operation, including the method of operation. [Note 3] The operating method described in Appendix 2, If the first topic falls under any of the pre-set important topics, the salesperson shall be presented with important topics other than those that fall under the first topic. The method of operation, which further includes the following. [Note 4] The operating method described in Appendix 2 or 3, To present to the salesperson important topics from among several pre-set important topics that do not fall under the third topic, The method of operation, which further includes the following. [Note 5] The operation method described in any of the appendices 2 to 4, If the first topic corresponds to a pre-defined important topic, the salesperson is presented with keywords that have been pre-associated with the important topic corresponding to the first topic but are not included in the dialogue. The method of operation, which further includes the following. [Note 6] The operation method described in any of the appendices 2 to 5, Each time the aforementioned first topic changes, Based on the audio data of the dialogue portion relating to the third topic, derive and remember the emotional states of the salesperson and the customer. The method of operation, which further includes the following. [Note 7] The operating method described in Appendix 6, The information processing device operates in a manner that derives and stores the emotional states of the salesperson and the customer based on the image data of the conversation. [Note 8] The operation method described in any of the appendices 2 to 7, Based on the above summary, present the salesperson with the next topic to be addressed. The method of operation, which further includes the following. [Note 9] The operation method described in any of the appendices 2 to 8, If the aforementioned first topic corresponds to a pre-defined trending topic, the salesperson shall be presented with a past summary or corresponding manual related to that trending topic. The method of operation, which further includes the following. [Note 10] The operation method described in any of the appendices 2 to 9, If the first topic does not change for a predetermined time or longer, it is determined that the first topic has changed. The method of operation, which further includes the following. [Note 11] An information processing device, We acquire audio data of conversations between sales staff and customers. Using speech recognition technology, the aforementioned speech data is converted into text data. Using co-occurrence analysis or topic modeling, extract the first topic, which is the current topic of the dialogue, from the text data. Each time a change in the first topic is detected, Identify the audio data of the dialogue portion related to the immediately preceding topic, the second topic. Based on the identified audio data, a summary of the dialogue portion is generated. The generated summary is presented to the salesperson. Control unit, An information processing device equipped with the following features. [Note 12] An information processing device, A control unit, based on audio data of a conversation between a salesperson and a customer, presents a summary of a past topic (the third topic) whenever the current topic (the first topic) of the conversation changes. An information processing device equipped with the following features. [Note 13] The information processing device described in Appendix 12, The control unit, An information processing device that, if the first topic corresponds to one of a set of important topics, presents the salesperson with important topics other than the one corresponding to the first topic. [Note 14] An information processing device as described in Appendix 12 or 13, The control unit, An information processing device that presents to the salesperson important topics from among several pre-set important topics that do not fall under the third topic. [Note 15] An information processing device described in any of Appendix 12 to 14, The control unit, If the first topic corresponds to a pre-defined important topic, an information processing device presents the salesperson with keywords that are pre-associated with the important topic corresponding to the first topic but are not included in the dialogue. [Note 16] An information processing device described in any of the appendices 12 to 15, The control unit, Each time the aforementioned first topic changes, An information processing device that derives and stores the emotional states of the salesperson and the customer based on audio data of the dialogue portion relating to the third topic. [Note 17] The information processing device described in Appendix 16, The control unit is an information processing device that derives and stores the emotional states of the salesperson and the customer based on the image data of the conversation. [Note 18] An information processing device described in any of the appendices 12 to 17, The control unit, An information processing device that, based on the summary above, presents the salesperson with the next topic to be addressed. [Note 19] An information processing device as described in any of the appendices 12 to 18, The control unit, An information processing device that, if the first topic corresponds to a pre-set trending topic, presents the salesperson with a past summary or corresponding manual related to that trending topic. [Note 20] An information processing device as described in any of the appendices 12 to 19, The control unit, An information processing device that determines that the first topic has changed if the first topic has not changed for a predetermined period of time or longer. [Explanation of symbols]

[0055] 1 System 10 Terminal devices 11 Communications Department 12 Output section 13 Input section 14 Storage section 15 Control Unit 20 Information Processing Devices 21 Communications Department 22 Memory section 23 Control Unit 30 Networks

Claims

1. A method for operating an information processing device, To acquire audio data of conversations between salespeople and customers, Using speech recognition technology, convert the aforementioned speech data into text data. Using co-occurrence analysis or topic modeling, derive the first topic, which is the current topic of the dialogue, from the text data. Each time a change in the first topic is detected, Identifying the audio data of the dialogue portion related to the immediately preceding topic, the second topic. Based on the identified audio data, generate a summary of the dialogue portion, and To present the generated summary to the salesperson, The method of operation, including the operation method.

2. A method for operating an information processing device, Based on audio data of conversations between salespeople and customers, whenever the current topic of the conversation (topic 1) changes, a summary of the past topic (topic 3) is presented. The method of operation, including the operation method.

3. The operating method according to claim 2, If the first topic falls under any of the pre-set important topics, the salesperson shall be presented with important topics other than those that fall under the first topic. The method of operation, which further includes the following.

4. The operating method according to claim 2, To present to the salesperson important topics from among several pre-set important topics that do not fall under the third topic, The method of operation, which further includes the following.

5. The operating method according to claim 2, If the first topic corresponds to a pre-defined important topic, the salesperson is presented with keywords that have been pre-associated with the important topic corresponding to the first topic but are not included in the dialogue. The method of operation, which further includes the following.

6. The operating method according to claim 2, Each time the first topic changes, Based on the audio data of the dialogue portion relating to the third topic, the emotional states of the salesperson and the customer are derived and stored. The method of operation, which further includes the following.

7. The operating method according to claim 6, The information processing device operates in a manner that derives and stores the emotional states of the salesperson and the customer based on the image data of the conversation.

8. The operating method according to claim 2, Based on the above summary, present the salesperson with the next topic to be addressed. The method of operation, which further includes the following.

9. The operating method according to claim 2, If the first topic mentioned above corresponds to a pre-defined trending topic, the salesperson shall be presented with a past summary or corresponding manual related to that trending topic. The method of operation, which further includes the following.

10. The operating method according to claim 2, If the first topic does not change for a predetermined time or longer, it is determined that the first topic has changed. The method of operation, which further includes the following.

11. An information processing device, We acquire audio data of conversations between sales staff and customers. Using speech recognition technology, the aforementioned speech data is converted into text data. Using co-occurrence analysis or topic modeling, extract the first topic, which is the current topic of the dialogue, from the text data. Each time a change in the first topic is detected, Identify the audio data of the dialogue portion related to the immediately preceding topic, the second topic. Based on the identified audio data, a summary of the dialogue portion is generated. The generated summary is presented to the salesperson. Control unit, An information processing device equipped with the following features.

12. An information processing device, A control unit, based on audio data of a conversation between a salesperson and a customer, presents a summary of a past topic (the third topic) whenever the current topic (the first topic) of the conversation changes. An information processing device equipped with the following features.

13. An information processing apparatus according to claim 12, The control unit, An information processing device that, if the first topic corresponds to one of a set of important topics, presents the salesperson with important topics other than the one corresponding to the first topic.

14. An information processing apparatus according to claim 12, The control unit, An information processing device that presents to the salesperson important topics from among several pre-set important topics that do not fall under the third topic.

15. An information processing apparatus according to claim 12, The control unit, If the first topic corresponds to a pre-set important topic, an information processing device presents the salesperson with keywords that are pre-associated with the important topic corresponding to the first topic but are not included in the dialogue.

16. An information processing apparatus according to claim 12, The control unit, Each time the first topic changes, An information processing device that derives and stores the emotional states of the salesperson and the customer based on audio data of the dialogue portion relating to the third topic.

17. An information processing apparatus according to claim 16, The control unit is an information processing device that derives and stores the emotional states of the salesperson and the customer based on the image data of the conversation.

18. An information processing apparatus according to claim 12, The control unit, An information processing device that, based on the summary above, presents the salesperson with the next topic to be addressed.

19. An information processing apparatus according to claim 12, The control unit, An information processing device that, if the first topic corresponds to a pre-set trending topic, presents the salesperson with a past summary or corresponding manual related to that trending topic.

20. An information processing apparatus according to claim 12, The control unit, An information processing device that determines that the first topic has changed if the first topic has not changed for a predetermined period of time or longer.

Images