WO2024075302A1

WO2024075302A1 - Information processing device, information processing method, and program

Info

Publication number: WO2024075302A1
Application number: PCT/JP2022/037722
Authority: WO
Inventors: 歩相名神山; 健一町田; 一比良松井
Original assignee: Ｎｔｔテクノクロス株式会社
Priority date: 2022-10-07
Filing date: 2022-10-07
Publication date: 2024-04-11

Abstract

An information processing device according to an embodiment of the present disclosure has: an extraction unit configured so that in response to a selection operation performed on a first display component expressing text in utterance units from a conversation between a plurality of people, a prescribed first character string is extracted from the text expressed by the first display component on which the selection operation was performed; and a search unit configured so as to use the extracted first character string as a search key and search knowledge expressing information for assisting with an inquiry response in the conversation.

Description

Information processing device, information processing method, and program

This disclosure relates to an information processing device, an information processing method, and a program.

Technology that supports operators in answering telephone calls at contact centers (also called call centers) has been known for some time (for example, Non-Patent Document 1). In such technology, a UI (user interface) called a response support screen is generally provided to the operator, and the operator can use various functions on this response support screen.

Functions available to operators on the above-mentioned customer support screen include the ability for the operator to search for knowledge such as FAQs (Frequently Asked Questions). There is also a function that automatically extracts keywords from the results of voice recognition during a call with a customer, searches for knowledge, and provides the operator with knowledge related to the customer's inquiry from among the knowledge obtained as a search result.

However, conventionally, operators have not been able to obtain appropriate knowledge in a short time. For example, when an operator searches for knowledge, the operator needs to copy keywords from the voice recognition results and enter them in a search area, and while appropriate knowledge can be obtained, it takes time to obtain the knowledge. Also, for example, when keywords are automatically extracted from the voice recognition results to search for and provide knowledge, the operator can obtain knowledge in a short time, but the knowledge may not be appropriate.

This disclosure has been made in light of the above, and aims to provide technology that allows appropriate knowledge to be obtained in a short period of time.

An information processing device according to one aspect of the present disclosure has an extraction unit configured to extract a predetermined first character string from the text represented by the first display component that is the target of a selection operation in response to a selection operation on the first display component that represents text in units of utterances in a conversation between multiple people, and a search unit configured to search for knowledge representing information for supporting response to inquiries in the conversation using the extracted first character string as a search key.

We provide technology that allows you to obtain appropriate knowledge in a short amount of time.

1 is a diagram illustrating an example of an overall configuration of a contact center system according to an embodiment of the present invention. FIG. 2 is a diagram illustrating an example of a functional configuration of the response support system and an operator terminal according to the present embodiment. FIG. 11 is a sequence diagram showing an example of a response support process according to the present embodiment. FIG. 13 is a diagram showing an example of a response support screen (part 1). FIG. 2 is a diagram showing an example of a response support screen (part 2). FIG. 3 is a diagram showing an example of a response support screen (part 3). 11 is a diagram for explaining an example of a relationship between a position where a selection operation is performed on an utterance part and an extracted keyword. FIG.

Below, one embodiment of the present invention will be described. In the following embodiment, a contact center system 1 will be described that is targeted at a contact center and enables contact center operators to obtain appropriate knowledge in a short time. However, the contact center is just one example and is not limited to this. The system can also be applied to other locations besides contact centers, such as offices and stores, where sales representatives or front desk staff of products or services obtain knowledge in the course of their work. More generally, the system can also be applied to cases where a person (which may be one person or multiple people) obtains knowledge during some kind of conversation between multiple people.

Furthermore, in the following description, we will assume that contact center operators handle customer inquiries and other tasks via voice calls, but this is not limited to this, and the same can be applied to tasks performed via text chat (including those that can send and receive stamps and attachments in addition to text), video calls, etc.

Here, knowledge refers to information necessary to answer an inquiry, or information that supports the answer. Typically, information obtained by searching FAQs, for example, falls under knowledge. Other examples of knowledge include information from manuals and web pages.

In recent years, as online business has accelerated and contact center operations have increased, the variety of inquiries has also increased, and the answers given by operators are also more diverse. Furthermore, at contact centers, operators are required to answer inquiries in a short time. For this reason, by introducing the contact center system 1 according to this embodiment and allowing operators to obtain appropriate knowledge in a short time (i.e., by obtaining useful knowledge for answering customer inquiries in a short time), it is expected that effects such as reducing the workload of operators and improving customer satisfaction can be achieved.

<Example of Overall Configuration of Contact Center System 1>
An example of the overall configuration of a contact center system 1 according to the present embodiment is shown in Fig. 1. As shown in Fig. 1, the contact center system 1 according to the present embodiment includes a response support system 10, a plurality of operator terminals 20, one or more supervisor terminals 30, a plurality of telephones 40, a PBX (Private Branch eXchange) 50, a NW switch 60, and a customer terminal 70. Here, the response support system 10, the operator terminals 20, the supervisor terminals 30, the telephones 40, the PBX 50, and the NW switch 60 are installed in a contact center environment E, which is a system environment of a contact center. Note that the contact center environment E is not limited to a system environment in the same building, and may be, for example, a system environment in a plurality of buildings that are geographically separated from each other.

The response support system 10 is a server or a group of servers that provides various functions (support functions) to support the operator in answering calls. There are various support functions provided by the response support system 10, but in this embodiment, at least the support functions shown in (1) and (2) below are provided.

(1) Speech text provision function: This function uses packets (voice packets) sent from the NW switch 60 to provide UI components that represent spoken text, which is a text of a voice call between an operator and a customer that has been converted using voice recognition technology with time information for each speaker. This function enables the operator to check the content of what is being said by the customer and the operator himself in real time during a call. Note that UI components are icons, buttons, figures, characters, etc. that are displayed on the screen. UI components may or may not be operable by the user. Below, UI components that represent spoken text are also referred to as "speech components."

(2) Knowledge search and provision function: This function searches for knowledge such as FAQs and provides the knowledge as a search result. This function allows an operator to check the information necessary to answer an inquiry, for example.

When an utterance part is selected by an operator, the response support system 10 extracts a keyword and a keyword related to that keyword (also called related keywords) from the utterance text represented by that utterance part, and provides a UI part for searching knowledge using the extracted keywords. Hereinafter, the UI part for searching knowledge using keywords is also called the "search part."

As a result, an operator can search for knowledge using keywords that correspond to a desired search component simply by selecting that search component. This makes it possible to search for knowledge in a short amount of time compared to, for example, manually copying keywords from spoken text and entering them in a search area. Furthermore, because not only keywords but also related keywords are extracted, an operator can obtain appropriate knowledge by selecting a search component that corresponds to the appropriate keyword.

The customer service support system 10 may be able to provide various support functions other than those described above in (1) and (2). For example, the customer service support system 10 may be able to provide all or some of the support functions described below in (3) to (14).

(3) Important matters provision function: This function provides the matters that the operator needs to explain to the customer or confirm (such matters are also called "important matters"). This function enables the operator to know what matters need to be explained to the customer and what matters need to be confirmed.

(4) Chat function: This function provides the ability to send and receive messages (so-called chat) between the operator and a supervisor, etc. This function enables the operator, for example, to make inquiries of the supervisor or ask for assistance. Note that a supervisor is, for example, a person who monitors the operator's calls and supports the operator's telephone response duties when a problem appears to be occurring or at the request of the operator.

(5) Support request function: This function is for requesting support from a supervisor, etc. This function enables an operator to request support from a supervisor, etc., for example, when a problem occurs or when a problem is likely to occur.

(6) Customer information provision function: This function provides information about the customer that the operator is currently serving (for example, attribute information such as name, age, sex, address, product and service purchase history, etc.). This function allows the operator to know various information about the customer that he or she is currently serving.

(7) Memo function: This function allows the operator to create memos of their choice. This function allows the operator to leave any content they wish as a memo (interaction record memo), for example.

(8) Previous call summary function: This function provides a summary of past calls of the customer that the operator is currently serving. This function allows the operator to check the summary text that summarizes the past calls of the customer.

(9) Call summary function: This function uses the text spoken during a call to create a summary text that summarizes the contents of the call. With this function, the operator can, for example, know the summary of the call after it has ended, or if the same customer calls again, check the summary of the previous call using the previous call summary provision function.

(10) Call analysis function: This function analyzes the text of speech during a call to identify the call reason and scene of the call, and whether the call is inbound or outbound. This function allows the operator to know the call reason and scene of the customer he or she is currently serving. In addition, for example, the important information provision function makes it possible to provide important information according to the call reason, scene, and inbound/outbound. Note that examples of scenes include the initial greeting (opening), confirmation of the inquiry, customer identity verification, response, and final greeting (closing).

(11) Call history storage function: This function stores, for example, information about the operator who answered the call, information about the customer of the call, the voice data of the call, the spoken text of the call, summary text, etc. as call history information. This function makes it possible to utilize the call history information for, for example, analyzing the quality of service and evaluating operators.

(12) Call history search function: This function searches for call history information stored by the call history storage function and provides the call history information as the search results to operators, supervisors, etc. This function enables operators, supervisors, etc. to search for call history information using desired search conditions and refer to the call history information obtained as the search results.

(13) Call monitoring function: This function provides information for supervisors to monitor operator calls (for example, the spoken text of the call and inappropriate words contained in the spoken text). This function enables supervisors to efficiently monitor the calls of the operators they are monitoring. Generally, one supervisor monitors the calls of several to a dozen operators.

(14) Analysis function: This function uses the call history information saved by the call history saving function to perform various analyses (e.g., analysis of response quality, etc.). This function enables those who wish to make calls with high response quality (e.g., response quality analysts and supervisors, etc.) to analyze response quality and consider various measures and strategies to improve it.

It goes without saying that the support functions shown in (3) to (14) above are all examples. In addition to these, the response support system 10 may be able to provide various support functions, for example, as described in the above-mentioned non-patent document 1.

The operator terminal 20 is a terminal such as a PC (personal computer) used by the operator.

The supervisor terminal 30 is a terminal such as a PC used by the supervisor.

Telephone 40 is an IP (Internet Protocol) telephone (such as a fixed IP telephone or a mobile IP telephone) used by an operator.

PBX 50 is a telephone exchange (IP-PBX) and is connected to a communications network 80 that includes a VoIP (Voice over Internet Protocol) network and a PSTN (Public Switched Telephone Network).

The NW switch 60 relays packets (voice packets) between the telephone 40 and the PBX 50, and also captures the packets and transmits them to the response support system 10.

The customer terminal 70 is a variety of terminals used by customers, such as smartphones, mobile phones, and landlines.

Note that the overall configuration of the contact center system 1 shown in FIG. 1 is an example, and is not limited to this, and other configurations may be used. For example, in the example shown in FIG. 1, the response support system 10 is included in the contact center environment E (i.e., the response support system 10 is an on-premise type), but all or part of the functions of the response support system 10 may be realized by a cloud service, etc. Similarly, in the example shown in FIG. 1, the PBX 50 is an on-premise type telephone exchange, but may be realized by a cloud service. Furthermore, if the operator terminal 20 has a telephone function, the contact center system 1 does not need to include a telephone set 40.

<Example of functional configuration of response support system 10 and operator terminal 20>
FIG. 2 shows an example of the functional configuration of the response support system 10 and the operator terminal 20 according to this embodiment.

<Response Support System 10>
As shown in FIG. 2, the customer service support system 10 according to the present embodiment includes a voice recognition unit 101 and a customer service support unit 102. These functional units are realized, for example, by a process in which one or more programs installed in the customer service support system 10 are executed by a processor such as a CPU (Central Processing Unit). The customer service support system 10 according to the present embodiment also includes a spoken text storage unit 103, a keyword storage unit 104, and a knowledge storage unit 105. These storage units are realized, for example, by a storage device such as a hard disk drive (HDD), a solid state drive (SSD), or a flash memory. However, at least one of these storage units may be realized, for example, by a storage device such as a database server connected to the customer service support system 10 via a communication network.

The speech recognition unit 101 performs speech recognition on the voice represented by the voice data contained in the voice packet received from the NW switch 60, and creates spoken text with time information for each speaker. The speech recognition unit 101 also stores these spoken texts in the spoken text storage unit 103 for each call. Note that, using existing speech recognition technology, it is possible to create spoken text with time information for each speaker from a voice that includes speech from multiple people.

The response support unit 102 provides the operator terminal 20 with a response support screen and UI parts displayed on the screen, as well as various support functions (i.e., various support functions including at least a spoken text provision function and a knowledge search and provision function). The response support screen and the UI parts displayed on the screen are represented by information such as HTML (Hypertext Markup Language), CSS (Cascading Style Sheets), JavaScript, etc.

In addition, when a speech part is selected on the response support screen, the response support unit 102 refers to the keyword information stored in the keyword storage unit 104 and extracts a keyword and related keywords from the speech text represented by that speech part. The response support unit 102 then displays a search part on the response support screen for searching knowledge using the extracted keyword.

The spoken text storage unit 103 stores, for each call, the spoken text with time information for each speaker in that call. In other words, for each call, the spoken text storage unit 103 stores, in chronological order, the spoken text for each speaker in that call.

The keyword storage unit 104 stores one or more pieces of keyword information. Keyword information is information in which a keyword is associated with a keyword related to it. For example, keyword information is expressed in the format of (keyword, related keyword set). Here, the related keyword set is a set of related keywords. The related keyword set may be an ordered set in which the related keywords are ordered in order of their relevance to the keyword, for example. Note that if there are no related keywords for a certain keyword, the related keyword set is an empty set.

An example of keyword information is (cashing, {contract, bank}). This keyword information indicates that the keyword is "cashing," and that the keywords related to "cashing" are "contract" and "bank."

The knowledge storage unit 105 stores one or more pieces of knowledge information. Knowledge information is information in which a keyword and knowledge related to that keyword are associated with each other. For example, knowledge information is expressed in the format of (keyword, knowledge). This makes it possible to search for knowledge related to a keyword using the keyword as a search key. However, the above format of knowledge information is only an example, and knowledge information may include various information other than keywords and knowledge, such as the number of times the knowledge has been referenced and evaluation information that evaluates the knowledge based on some criteria. In addition, evaluation information may include, for example, the average value of the operator's evaluation of the knowledge using a graded numerical value (for example, the average value of a five-point evaluation), the total number of times the operator has evaluated the knowledge as appropriate, such as "Like" or "GOOD", etc.

Note that although not shown in the example shown in FIG. 2, the response support system 10 also has various storage units and other functional units that are used to realize various support functions. For example, it has an operator information storage unit in which operator information including each operator's operator ID and attributes (name, affiliated organization, etc.) is stored, a call history information storage unit in which each call history information is stored, etc.

<Operator terminal 20>
2, the operator terminal 20 according to the present embodiment has a UI control unit 201. The UI control unit 201 is realized, for example, by a process in which one or more programs (e.g., a web browser, etc.) installed in the operator terminal 20 are executed by a processor such as a CPU.

The UI control unit 201 displays the response support screen provided by the response support system 10 on the display, and dynamically (in real time) displays each speech part received from the response support system 10 on the response support screen. Also, when the UI control unit 201 receives knowledge from the response support system 10, it displays the knowledge on the response support screen. Furthermore, the UI control unit 201 accepts various input operations by the operator on the response support screen (for example, pressing a button, inputting characters, etc.).

<Response support processing>
Hereinafter, assuming that a voice call is being made between a customer and an operator, the response support process for supporting the operator in answering the telephone call will be described with reference to Fig. 3. In the following, it is assumed that a response support screen is displayed on the display of the operator terminal 20 used by the operator by the UI control unit 201. Note that while a voice call is being made between a customer and an operator, voice packets transmitted and received between the customer terminal 70 used by the customer and the operator terminal 20 used by the operator are transmitted (transferred) to the response support system 10 by the NW switch 60.

The voice recognition unit 101 of the response support system 10 performs voice recognition on the voice represented by the voice data contained in the voice packet received from the NW switch 60, and creates a spoken text with time information for each speaker (step S101).

Next, the voice recognition unit 101 of the response support system 10 stores (memorizes) the speech text created in step S101 above in the speech text storage unit 103 (step S102).

Next, the response support unit 102 of the response support system 10 transmits the speech component representing the speech text created in step S101 above to the operator terminal 20 (step S103).

When the UI control unit 201 of the operator terminal 20 receives the speech parts from the customer service support system 10, it displays the speech parts on the customer service support screen (step S104).

Here, an example of the customer service support screen displayed by the UI control unit 201 is shown in FIG. 4. The customer service support screen 1000 shown in FIG. 4 includes an utterance content display field 1100. In the utterance content display field 1100, speech parts received from the customer service support system 10 are displayed in chronological order for each speaker. For example, in the example shown in FIG. 4, speech parts 1111-1113 representing the operator's utterance text and speech parts 1121-1123 representing the customer's utterance text are displayed in chronological order. This allows the operator to check his or her own utterance content and the customer's utterance content in chronological order.

In addition, each speech part displayed in the speech content display field 1100 may be associated with a UI part (hereinafter also referred to as a speech evaluation part) for evaluating that the speech text represented by that speech part is useful for narrowing down knowledge during knowledge search. This speech evaluation part may be called, for example, a "Like" button or a "GOOD" button related to the speech text. In the example shown in FIG. 4, speech parts 1111-1113 are associated with speech evaluation parts 1131-1133, respectively, and similarly, speech parts 1121-1123 are associated with speech evaluation parts 1141-1143. The operator can perform a selection operation on the speech evaluation parts displayed in the speech content display field 1100 on the response support screen 1000 shown in FIG. 4. For example, the operator can use an input device such as a pointing device, touch panel, or keyboard to perform a selection operation to select a desired utterance evaluation component from among utterance evaluation components 1131 to 1133 and 1141 to 1143. The example shown in FIG. 4 shows a case where a selection operation is performed on utterance evaluation component 1141.

This allows the operator to evaluate spoken text that is useful for narrowing down knowledge when searching for knowledge, and as described below, it becomes possible to narrow down knowledge using keywords contained in this spoken text. Note that in the example shown in Figure 4, the spoken text is evaluated as to whether it is useful for narrowing down knowledge when searching for knowledge, but this is not limited to this, and it may be evaluated, for example, using a graded numerical value, etc.

Here, the operator can perform a selection operation on the speech part displayed in the speech content display field 1100 on the customer service support screen 1000 shown in FIG. 4. For example, the operator can use an input device such as a pointing device, touch panel, or keyboard to perform a selection operation to select a desired speech part from among the speech parts 1111-1113 and 1121-1123 displayed in the speech content display field 1100. In the following, the explanation will be continued assuming that a selection operation has been performed on a certain speech part displayed in the speech content display field 1100 on the customer service support screen 1000 shown in FIG. 4.

In the following, it is assumed that one keyword is included in the speech text represented by the speech part. However, the speech text represented by the speech part may not contain any keyword, or may contain multiple keywords. If the speech text does not contain a keyword, no keyword is extracted in step S106 described below. On the other hand, if the speech text contains multiple keywords, in step S106 described below, for example, all of the multiple keywords may be extracted, or each time a selection operation is performed on the speech part, the keywords may be extracted in order starting from the keyword that first appears in the speech text.

When a selection operation is performed on the speech part, the UI control unit 201 of the operator terminal 20 transmits operation information indicating that the speech part has been selected to the customer service support system 10 (step S105). Note that the operation information includes, for example, identification information (e.g., a part ID, etc.) of the speech part for which the selection operation has been performed. However, the UI control unit 201 of the operator terminal 20 may transmit, for example, the speech text represented by the speech part to the customer service support system 10 instead of the operation information, or may transmit identification information (e.g., a speech text ID, etc.) of the speech text represented by the speech part.

When the response support unit 102 of the response support system 10 receives operation information from the operator terminal 20, it extracts keywords (including related keywords) from the spoken text represented by the speech component of the component ID included in the operation information (step S106). That is, the response support unit 102 refers to the keyword information stored in the keyword storage unit 104 to extract keywords from the spoken text, and also extracts keywords related to the keywords. Note that a collection of keywords and their related keywords (related keywords) may be called a "preset" or a "keyword preset", etc.

For example, assume that keyword information (cashing, {contract, bank}) etc. is stored in the keyword storage unit 104, and a selection operation is performed on the speech component 1122 displayed in the speech content display field 1100 on the customer service support screen 1000 shown in FIG. 4. In this case, the customer service support unit 102 extracts the keyword "cashing" from the speech text "I'd like to confirm something about that cashing service" represented by the speech component 1122, and also extracts the keywords "contract" and "bank" related to the keyword "cashing".

Next, the customer service support unit 102 of the customer service support system 10 transmits a search component for searching knowledge using the keywords extracted in step S106 above to the operator terminal 20 (step S107).

For example, if the keywords extracted in step S106 above are "cashing," "contract," and "bank," a search component for searching knowledge for "cashing," a search component for searching knowledge for "contract," and a search component for searching knowledge for "bank" are transmitted to the operator terminal 20.

When the UI control unit 201 of the operator terminal 20 receives the search components from the customer service support system 10, it displays these search components on the customer service support screen (step S108). These search components are displayed, for example, in the search component display field 1200 of the customer service support screen 1000 shown in FIG. 4. At this time, for example, the search component corresponding to the keyword extracted from the spoken text is displayed on the leftmost side (the display position with the highest priority). In addition, the search components corresponding to the related keywords may be displayed, for example, from the left in order of relevance to the keyword, or from the left in order of the related keywords being extracted. For example, in the example shown in FIG. 4, the search component 1210 corresponding to the keyword "cashing" is displayed on the leftmost side (the display position with the highest priority), followed by the search component 1220 corresponding to the related keyword "contract" (the display position with the second highest priority), and then the search component 1230 corresponding to the related keyword "bank" (the display position with the third highest priority). Note that it is considered that the higher the display position of the search component is, the more likely it is that appropriate knowledge will be obtained when searching for knowledge with the keyword corresponding to that search component.

Note that the display mode (e.g., color of UI components) of the search component corresponding to a keyword and the search component corresponding to a related keyword may be different. The example shown in FIG. 4 shows a case where the display mode of the search component 1210 corresponding to a keyword is different from that of the search components 1220 to 1230 corresponding to related keywords.

Here, the operator can perform a selection operation on the search component displayed in the search component display field 1200 on the customer service support screen 1000 shown in FIG. 4. For example, the operator can use an input device such as a pointing device, touch panel, or keyboard to perform a selection operation to select a desired search component from among the search components 1210 to 1230 displayed in the search component display field 1200. In the following, the explanation will be continued assuming that a selection operation has been performed on a certain search component displayed in the search component display field 1200 on the customer service support screen 1000 shown in FIG. 4.

When a selection operation is performed on the search component, the UI control unit 201 of the operator terminal 20 transmits operation information indicating that the search component has been selected to the customer service support system 10 (step S109). Note that the operation information includes, for example, identification information (e.g., a component ID, etc.) of the search component on which the selection operation has been performed. However, the UI control unit 201 of the operator terminal 20 may transmit, for example, a keyword corresponding to the search component to the customer service support system 10 instead of the operation information, or may transmit identification information (e.g., a keyword ID, etc.) of the keyword corresponding to the search component.

When the customer service support unit 102 of the customer service support system 10 receives operation information from the operator terminal 20, it searches for knowledge using a keyword corresponding to the search component of the component ID included in the operation information (step S110). That is, the customer service support unit 102 searches for knowledge information stored in the knowledge storage unit 105 using the keyword as a search key, and obtains knowledge information related to the keyword as the search result.

For example, when an operator performs a selection operation on search component 1210, the knowledge information stored in knowledge storage unit 105 is searched for using the keyword "caching" corresponding to this search component 1210 as a search key. This allows knowledge information related to the keyword "caching" to be searched for. Note that a known method may be used as the knowledge search method.

Here, if an utterance evaluation part is selected by the operator, in step S110, a knowledge search is performed using the keywords contained in the speech text associated with that utterance evaluation part as an AND condition. For example, if the utterance evaluation part 1141 is selected on the response support screen 1000 shown in FIG. 4, a knowledge search is performed using the keyword "ABC card" contained in the speech text associated with this utterance evaluation part 1141 as an AND condition. Specifically, for example, if the keyword corresponding to the search part selected by the operator is "cashing," knowledge information is searched for using "cashing" and "ABC card" as search keys. In this way, when an utterance evaluation part is selected by the operator, it is possible to narrow down knowledge during a knowledge search by the keywords contained in the speech text associated with that utterance evaluation part.

Next, the customer service support unit 102 of the customer service support system 10 transmits the knowledge contained in the knowledge information searched in step S110 and associated information to the operator terminal 20 as a search result (step S111). Note that examples of information associated with knowledge include the number of times the knowledge has been referenced, evaluation information for the knowledge, etc.

When the UI control unit 201 of the operator terminal 20 receives the search results from the customer service support system 10, it displays the search results on the customer service support screen (step S112). The search results are displayed, for example, in the search result display field 1300 of the customer service support screen 1000 shown in FIG. 4. In the search result display field 1300 shown in FIG. 4, knowledge items 1301 to 1302 are displayed as search results. Note that by selecting the desired knowledge from the knowledge items displayed in the search result display field 1300, more detailed information regarding the selected knowledge item may be displayed.

At this time, in the search result display area 1300, each piece of knowledge obtained as a search result may be displayed in a display order according to information accompanying the knowledge. For example, each piece of knowledge may be displayed in order of the number of times the knowledge has been referenced, or in order of the best evaluation represented by the evaluation information (for example, in order of the greatest total number of "Likes" or "GOOD" votes, or in order of the highest average value of a graded numerical evaluation, etc.).

The knowledge displayed in the search result display field 1300 may be associated with a UI component (hereinafter also referred to as a knowledge evaluation component) for evaluating whether the knowledge is useful for answering an inquiry. This knowledge evaluation component may be referred to as, for example, a "Like" button or a "GOOD" button related to the knowledge. In the example shown in FIG. 4, knowledge 1301-1302 are associated with knowledge evaluation components 1311-1312, respectively. The operator can perform a selection operation on the knowledge evaluation components displayed in the search result display field 1300 on the response support screen 1000 shown in FIG. 4. For example, the operator can perform a selection operation to select a desired knowledge evaluation component from the knowledge evaluation components 1311-1312 using an input device such as a pointing device, touch panel, or keyboard. When a selection operation is performed on a knowledge evaluation component, operation information indicating that the knowledge evaluation component has been selected is sent to the response support system 10, and as a result, the evaluation information of the knowledge information stored in the knowledge storage unit 105 that includes the knowledge associated with that knowledge evaluation component is updated.

This makes it possible to display knowledge obtained as search results that is more useful in answering the inquiry (i.e., knowledge of higher quality) higher in the display order, and less useful knowledge lower in the display order.

As described above, in the contact center system 1 according to this embodiment, the operator can display a search component on the customer support screen for searching for knowledge using keywords contained in the spoken text or related keywords, simply by performing a selection operation on a speech component that represents the spoken text of a customer or the operator himself/herself. The operator can then search for knowledge using the keywords that correspond to the desired search component, simply by performing a selection operation on the customer support screen. This allows the operator to search for knowledge in a short amount of time. Furthermore, search components that correspond to related keywords as well as keywords are displayed, and the operator can select a search component that corresponds to the appropriate keyword from among these search components, thereby enabling the operator to obtain appropriate knowledge.

Therefore, according to the contact center system 1 of this embodiment, when responding to inquiries from customers, the operator can provide appropriate answers without making the customer wait, which, for example, can result in improved customer satisfaction.

As in the past, the operator can also search for knowledge related to a desired keyword by entering the desired keyword in the search area 1400 on the response support screen 1000 shown in FIG. 4 and pressing the search button 1500.

<Modification>
A modification of this embodiment will now be described.

<Modification 1>
In the above embodiment, knowledge is searched when a selection operation is performed on a search component, but for example, a knowledge search may be performed in advance for keywords (including related keywords) extracted from utterance components, and when the mouse is placed over a search component, the search results for the keywords corresponding to that search component (or the top several pieces of knowledge included in the search results) may be displayed in a pop-up. For example, when the mouse is placed over a search component 1210 on the customer service support screen 1000 shown in FIG. 5, the search results for the keywords corresponding to that search component 1210 may be displayed on a pop-up screen 2100.

This allows operators to simply hover the mouse over a search component and easily check what kind of search results will be obtained when performing a knowledge search using the keywords that correspond to that search component.

<<Modification 2>>
In the above embodiment, depending on the keyword extracted from the speech text represented by the speech part, appropriate knowledge may not be obtained. For example, when the keyword extracted from the speech text is an abstract term or a term with multiple meanings, a lot of knowledge may be obtained as a search result, and the narrowing down of knowledge may be insufficient. For this reason, when the keyword extracted from the speech text represented by the speech part is a specific keyword (for example, a keyword representing an abstract term or a term with multiple meanings), a suggestion (proposal or suggestion) for further narrowing down the knowledge during the knowledge search may be presented to the operator.

For example, in the response support screen 1000 shown in FIG. 6, a suggestion 3100 for further narrowing down knowledge is presented to the operator when searching for knowledge using the keyword "caching" corresponding to the search component 1210. This suggestion 3100 displays the search results when searching for knowledge by further narrowing down knowledge related to the keyword "caching".

This allows the operator to obtain appropriate knowledge even if the keywords extracted from the spoken text represented by the speech components represent abstract terms or terms with multiple meanings.

<Modification 3>
If multiple keywords are contained in the speech text represented by the speech part, in the above step S106, the keyword located closest to the position (operation position) where the selection operation was performed on the speech part may be extracted depending on the position (operation position) where the selection operation was performed.

For example, assume that a selection operation is performed on speech component 4100 shown in FIG. 7. Here, assume that the speech text represented by speech component 4100 contains keyword 4101 representing "ABC service" and keyword 4102 representing "XX function." In this case, for example, if a selection operation is performed on operation position 4103, keyword 4101 is extracted in step S106 above. On the other hand, for example, if a selection operation is performed on operation position 4104, keyword 4102 is extracted in step S106 above.

This allows the operator to extract only the desired keywords from the speech text represented by the speech component depending on their operating position.

<Modification 4>
In step S106 above, keywords and related keywords are extracted, but for example, only the keywords may be extracted without extracting the related keywords. Also, when only the keywords are extracted, knowledge may be automatically searched for using the keywords and the search results may be presented to the operator. This allows the operator to search for knowledge using keywords contained in the utterance text represented by the utterance parts simply by performing a selection operation on the utterance parts, making it possible to search for knowledge in a shorter time.

<Modification 5>
When displaying search components corresponding to related keywords on the customer support screen, the priority (display order) of those search components may be determined according to their relationship with the topic of the conversation between the customer and the operator. For example, after determining the topic of the conversation between the customer and the operator, the likelihood (e.g., probability) that the related keyword is the topic may be calculated, and the priority (display order) of the search components corresponding to the related keywords may be determined in descending order of likelihood. Note that the topic may be determined using a known topic determination technique.

<Modification 6>
The search components displayed on the customer service support screen may be deleted, for example, each time a selection operation is performed on an utterance component, or may be added each time a selection operation is performed on an utterance component. In addition, if a search component is added each time a selection operation is performed on an utterance component, the oldest search component may be deleted in order when the maximum number of search components that can be displayed on the customer service support screen is exceeded.

The present invention is not limited to the specifically disclosed embodiments above, and various modifications, changes, and combinations with known technologies are possible without departing from the scope of the claims.

Reference Signs List 1 Contact center system 10 Response support system 20 Operator terminal 30 Supervisor terminal 40 Telephone 50 PBX
60 NW switch 70 Customer terminal 80 Communication network 101 Voice recognition unit 102 Response support unit 103 Speech text storage unit 104 Keyword storage unit 105 Knowledge storage unit 201 UI control unit E Contact center environment

Claims

an extraction unit configured to extract, in response to a selection operation on a first display part that represents text in units of utterances in a conversation between multiple people, a predetermined first character string from the text represented by the first display part that is the target of the selection operation;
a search unit configured to search for knowledge representing information for supporting a response to an inquiry in the conversation, using the extracted first character string as a search key;
An information processing device having the above configuration.
The extraction unit is
configured to extract the first string and a second string related to the first string;
The search unit is
2. The information processing device according to claim 1, configured to search the knowledge using the first character string or the second character string corresponding to the second display part that is the target of the selection operation, in response to a selection operation on a second display part for performing the search using the first character string or the second character string as a search key.
The second display component corresponding to the second character string is
3 . The information processing device according to claim 2 , wherein the second character string is displayed on a UI in order of relevance between the second character string and the first character string, or in order of likelihood that a topic of the second character string is the same as a topic of the conversation.
The extraction unit is
configured to extract the first string and a second string related to the first string;
The search unit is
performing the search using each of the first character string and the second character string as a search key;
2. The information processing device according to claim 1, configured to display, in response to a superimposition operation of a pointer position on a second display component for performing the search using the first character string or the second character string as a search key, knowledge obtained as a search result when the search is performed using the first character string or the second character string corresponding to the second display component that is the target of the superimposition operation as a search key, or a predetermined number of pieces of knowledge included in the search result, on a UI.
The search unit is
5. The information processing device according to claim 1, further comprising: a first display part that is evaluated by a user as representing a useful utterance among the first display parts, and a predetermined character string contained in the text displayed by the first display part is also used as a search key to search the knowledge.
The search unit is
The information processing apparatus according to claim 1 , wherein the knowledge obtained as a search result is displayed on the UI in an order based on evaluation information for the knowledge.
The extraction unit is
2. The information processing device according to claim 1, further configured to extract the first character string located closest to the position where the selection operation was performed when a plurality of the first character strings are included in the text represented by the first display component that was the target of the selection operation.
an extraction step of extracting a predetermined first character string from the text represented by a first display part that is the target of a selection operation, the first display part representing text in units of utterances in a conversation between multiple people;
a search procedure for searching for knowledge representing information for supporting a response to an inquiry in the conversation, using the extracted first character string as a search key;
An information processing method implemented by a computer.
an extraction step of extracting a predetermined first character string from the text represented by a first display part that is the target of a selection operation, the first display part representing text in units of utterances in a conversation between multiple people;
a search procedure for searching for knowledge representing information for supporting a response to an inquiry in the conversation, using the extracted first character string as a search key;
A program that causes a computer to execute the following.