WO2011063761A1

WO2011063761A1 - Method, device and system for speech detection in call center

Info

Publication number: WO2011063761A1
Application number: PCT/CN2010/079207
Authority: WO
Inventors: 王剑锋
Original assignee: 华为技术有限公司
Priority date: 2009-11-27
Filing date: 2010-11-27
Publication date: 2011-06-03
Also published as: CN102082879B; CN102082879A

Abstract

The embodiments of the present invention provide a method, device and system for speech detection in a call center. The method includes: during a conversation process between an operator and a user, firstly, a call center platform initiates a call conversation recording by a recording device, and transmits a speech media stream, which is generated in real-time, of the operator to a speech detection server, so that the speech detection server generates an event message in real time according to the received speech media stream; and then the call center platform receives the event message generated by the speech detection server, and issues a corresponding event prompt to the operator and/or a quality inspector. Thus monitoring the conversation content in real-time can be realized. Thereby, the service quality of the real-time conversation is improved, the work improvement of the operator is promoted and the service quality of the call center is improved.

Description

Call center voice detection method, device and system

This application claims the priority of the Chinese Patent Application entitled "Call Center Voice Detection Method, Apparatus and System" submitted by the Chinese Patent Office on November 27, 2009, with the application number 200910237999.0. The content of the prior application document is approved. The citations are incorporated herein by reference.

Technical field

The present invention relates to the field of network communications, and in particular, to a method, device and system for voice detection in a call center. Background technique

At present, in the customer service of the call center, the primary attendant often needs the monitoring operation of the quality inspector or the squad leader to assist in the real-time call, and the recording can be provided for the quality inspection afterwards. With the maturity of speech recognition technology, the existing methods are generally combined with the techniques of voice emotion detection and keyword detection to improve the ability of agent services.

In the prior art, the quality inspector often processes the call recording file generated by the agent by the voice detection server, including the emotion detection server and/or the keyword detection server, to identify the call recording in the call. Abnormal emotional points and keywords, and deal with them accordingly. Summary of the invention

The embodiment of the invention provides a method, a device and a system for voice detection of a call center, which can monitor the content of the call in real time to improve the service quality of the real-time call, thereby promoting the improvement of the operator's work, and timely informing the corresponding Quality inspectors, which in turn improve the quality of service in the call center.

The embodiment of the present invention provides a method for voice detection of a call center, where the method includes: during a call between an operator and a user, the call center platform initiates a call call recording through a recording device, and generates a real-time generated caller for the attendant. The voice media stream is transmitted to the voice detection server, so that the voice detection server generates an event in real time according to the received voice media stream. Message

The call center platform receives the event message generated by the voice detection server, and sends a corresponding event prompt to the operator and/or the quality inspector.

The embodiment of the present invention further provides a method for voice detection of a call center, where the method includes: receiving, during a call between the operator and the user, a voice media stream sent by the call center platform for the operator in real time;

Generating emotional points and/or keywords in real time according to the voice media stream, and generating corresponding event messages according to the detected emotional points and/or keywords;

The generated event message is sent to the call center platform.

The embodiment of the invention further provides a device for voice detection of a call center, comprising:

a call call recording activation unit, configured to initiate a call call recording by the recording device during the conversation between the operator and the user, and transmit the voice media stream generated for the attendant in real time to the voice detection server, so as to facilitate the voice detection The server generates an event message in real time according to the received voice media stream;

a recording device, configured to initiate a call call recording according to the request of the call call recording activation unit, and generate a voice media stream for the attendant in real time;

The event message processing unit is configured to receive an event message generated by the voice detection server, and issue a corresponding event prompt to the operator and/or the quality inspector.

a voice media stream receiving unit, configured to receive, during a call between the operator and the user, a voice media stream sent by the call center platform for the attendant in real time;

An event message generating unit, configured to detect an emotional point and/or a keyword in real time according to the received voice media stream, and generate a corresponding event message according to the detected emotional point and/or keyword;

An event message sending unit is configured to send the generated event message to the call center platform. The embodiment of the invention further provides a system for call center voice detection, the system comprising a call center platform and a voice detection server, wherein: The call center platform is configured to initiate a call call recording by the recording device during the call between the operator and the user, and transmit the voice media stream generated for the attendant in real time to the voice detection server; and receive the voice detection The event message generated by the server, and the corresponding event prompt is sent to the operator and/or the quality inspector;

The voice detection server is configured to receive, during a call between the operator and the user, a voice media stream sent by the call center platform for the attendant in real time; and detect an emotional point and/or a key according to the received voice media stream in real time. Words, and generating corresponding event messages according to the detected emotional points and/or keywords, and transmitting the generated event messages to the call center platform.

It can be seen from the technical solution provided above that during the conversation between the operator and the user, the call center platform first initiates the call call recording through the recording device, and transmits the voice media stream generated for the attendant in real time to the voice detection. a server, so that the voice detection server generates an event message in real time according to the received voice media stream; then the call center platform receives the event message generated by the voice detection server, and sends a corresponding event prompt to the operator and/or Or quality inspector. In this way, real-time monitoring of the call content can be realized, thereby improving the service quality of the real-time call, and promoting the work improvement of the attendant, thereby improving the service quality of the call center. DRAWINGS

FIG. 1 is a schematic flowchart of a call center voice detection method according to an embodiment of the present invention; FIG. 2 is a schematic flow chart of another call center voice detection method according to an embodiment of the present invention;

3 is a schematic flowchart of a call center voice detection method according to an embodiment of the present invention; FIG. 4 is a schematic diagram of a network structure for implementing call center voice detection by using a queuing machine according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of signaling interaction performed by a queuing machine to implement call center voice detection according to an embodiment of the present invention; FIG.

6 is a schematic diagram of an interface for setting an emotion detection quality inspection task in advance in a specific application example; 7 is a schematic diagram of displaying on the agent interface and the monitoring interface of the quality inspector after the real-time detection by the emotion detection server in a specific application example;

FIG. 8 is a schematic structural diagram of a call center voice detecting apparatus according to an embodiment of the present invention; FIG. 9 is another schematic structural diagram of a call center voice detecting apparatus according to an embodiment of the present invention; A schematic diagram of the structure of a call center voice detecting device;

FIG. 11 is a schematic structural diagram of a call center voice detection system according to an embodiment of the present invention. detailed description

The embodiment of the invention provides a method, a device and a system for voice detection of a call center. By modifying the call center platform, real-time keywords and/or emotional prompts can be obtained during the agent call to improve the attendant's The quality of service, and the quality inspector can also perform corresponding processing operations through real-time information. The above call center platform can be a Computer Telephony Integration Platform (CTI).

FIG. 1 is a schematic flowchart of a method provided by an embodiment of the present invention, where the method includes: Step 11: During a call between an attendant and a user call, the call center platform initiates a call call recording through a recording device, and generates a real-time call. The voice media stream for the attendant is transmitted to the voice detection server, so that the voice detection server generates an event message in real time according to the received voice media stream.

Step 12: The call center platform receives an event message generated by the voice detection server, and sends a corresponding event prompt to the operator and/or the quality inspector.

Through the implementation of the above technical solutions, real-time monitoring of the call content can be realized through the call center platform, thereby improving the service quality of the real-time call, and promoting the work improvement of the attendant, thereby improving the service quality of the call center.

FIG. 2 is a schematic diagram of another call center voice detection method according to an embodiment of the present invention, where the method includes: Step 21: Receive a voice media stream sent by the call center platform for the attendant in real time during the conversation between the operator and the user.

Step 22: detecting an emotional point and/or a keyword according to the voice media stream in real time, and generating a corresponding event message according to the detected emotional point and/or keyword.

Step 23: Send the generated event message to the call center platform.

Through the implementation of the above technical solutions, real-time monitoring of the call content can be realized through the call center platform, thereby improving the service quality of the real-time call, and promoting the work improvement of the operator, thereby improving the service quality of the call center.

For a better description of the method embodiments of the present invention, the specific embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 3 is a schematic flowchart of the method provided by the embodiment of the present invention, the method includes:

Step 31: After the attendant starts a call with the user, the call center platform initiates a call call recording through the recording device, and the recording device generates a voice media stream for the attendant in real time, and transmits the real-time generated voice media stream for the attendant. Give the voice detection server.

Specifically, first, a call is accessed, and the attendant starts a call with the user. During the call between the attendant and the user call, the call center platform can initiate a call call recording through the recording device. The above voice detection server may generally include an emotion detection server and/or a keyword detection server. Specifically, only the emotion detection server, or only the keyword detection server, or both may exist according to actual needs.

In a specific implementation process, the process of the call center platform initiating the call call recording by the recording device may be: First, the call center platform requests the recording device to perform voice detection on the call, and then the recording device can initiate a call call according to the request. Recording, and real-time generation of a voice media stream for the attendant.

In addition, in the actual application, the quality inspection task of the voice detection may be preset, so that the call center platform requests the recording device to perform voice detection according to the preset quality inspection task. For example, the quality inspection task that can be preset includes one or more of the following: The quality inspection of the specified time period and the quality inspection of the designated operator. If the quality check task of the specified call is set, when the accessed call meets the specified call condition, the call center platform is triggered to request the voice recording device to perform voice detection on the call; if the designated operator's quality inspection task is set, the access is performed. When the call is made to a call with the designated operator, the call center platform is triggered to request the voice recording device to perform voice detection on the call. The above-mentioned preset quality inspection tasks can be set autonomously by the operator in the call center system.

It should be noted that, in the specific implementation process, the foregoing recording device may include various types of recording devices supporting the call center, for example, as a queuing machine of a call center access device, etc., or may be a local agent recording device or an IP packet capture recording. Device, etc. For example, the local agent recording device installs the corresponding call center recording control software on the local computer of each agent. When the call center platform requests to record a certain agent call, the local agent recording device is on the local agent. The call is recorded and converted into a media stream or a local file in real time; and the above IP capture recording device captures the IP image through the switch or router through which the voice IP stream passes, and it applies to each IP device through the network device. The voice packet is analyzed, and the IP packet stream of the specific IP is intercepted and summarized into a voice stream or file. When the call center platform requests to record a certain agent call, the IP packet capture recording device can intercept the corresponding report according to the request. Wen came to form the corresponding recording file. In addition to the recording apparatus mentioned above, other ways in which call recording can be realized can be applied to the present invention.

Step 32: After the voice detection server receives the voice media stream generated for the attendant in real time, the voice detection server detects the emotion point and/or the keyword in real time according to the received voice media stream, and according to the detected emotion Points and/or keywords to generate corresponding event messages.

Specifically, if the voice detection server detects a preset emotion point and/or keyword, an event message for the emotion point and/or the keyword is generated. For example, if the emotion point is set to “anger” in advance, When the voice detection server (in this case, the emotion detection server) detects an "angry" emotion in the received voice media stream, an "angry" event message can be generated.

Similarly, the corresponding keyword can also be set in advance, and the keyword detection server performs the check. The event message of the corresponding keyword may be generated when the corresponding keyword is detected in the received voice media stream. In addition, if no pre-set mood points and/or keywords are detected, no event message may be generated or an event message with a null result may be generated.

Step 33: The call center platform receives the event message generated by the voice detection server, and may send a corresponding event prompt to the attendant and/or the quality inspector, and the attendant and/or the quality inspector perform corresponding according to the event prompt. Processing.

In addition, in a specific implementation process, after the corresponding event prompt is sent to the attendant and/or the quality inspector, the service system of the call center platform may further perform a corresponding service operation prompt according to the event message, for example, an interface. Menu changes, termination of calls or voice reduction operations; or the quality inspector can perform corresponding business processing according to the received event prompts, such as inserting a call or forcing idle to ensure better call quality.

The following is a specific example to illustrate the process of supporting a recording device of a call center, such as a queuing machine, to implement a call recording process. As shown in FIG. 4, the call is implemented by a queuing machine according to an embodiment of the present invention. Schematic diagram of the network structure of the central voice detection, in which: the call center platform (CTI platform in this specific example) generates the voice media stream of the current agent attendant to the voice detection server in real time through the site channel recording mechanism of the queue machine; The detection server performs real-time detection on the emotions and keywords preset by the operator, and generates corresponding event messages to be sent to the CTI platform through the queuing machine; the CTI platform provides the corresponding event prompts to the current agent attendant and the quality inspector. Schematic diagram of signaling interaction of voice detection. Before accessing the call, the quality inspector can first set the corresponding voice detection server, that is, set the quality inspection task of emotion detection and/or keyword detection. The setting of the quality inspection task can refer to the call center. A general method, such as specifying a call, specifying a time period, or specifying an attendant. The specific example here sets a quality inspection task that specifies the operator's emotion detection and/or keyword detection. Although emotion detection and keyword detection belong to completely different voice detection servers, they can be unified for the process of CTI and queue machine. The difference in group network is mainly emotion detection server and key. Word detection server replacement.

Specifically, if the quality inspector sets the quality inspection task of the emotion detection, the key emotions for the operator, such as anger, happiness, etc., can be set, and when the emotion detection server detects the corresponding emotion in the operator's call, it will be generated. Corresponding to the emotional event message; if the quality inspector sets the quality inspection task for keyword detection, for example, setting keywords for the caller's call, polite expressions such as "hello" and "thank you" or other restrictions, as the keyword detection server When the set keyword is displayed during the attendant call, a corresponding keyword event message is generated.

After the above settings, when the call is accessed, it can be operated according to the sequence flow shown in Figure 5, including:

501. A calling user makes a voice call to the call center, and the voice call is first connected to the queue.

502. The queuing machine receives the voice call, and applies for a call route to the CTI.

503. The CTI allocates the call to an attendant according to a routing algorithm, and establishes a voice connection between the calling user and the operator assigned the corresponding agent.

504. The operator calls the host user to make a call.

505. According to the configuration policy, if the CTI is found to meet the conditions for starting the voice detection quality check task, the CTI requests voice detection for the call.

506. After receiving the real-time voice detection application of the CTI, the queuing machine starts the call recording, performs real-time recording on the call, and generates a media stream to be sent to the voice detection server.

507. The voice detection server performs detection and identification of a key event on the real-time media stream. When the set mood and/or keyword is detected, the event message of the sentiment and/or keyword detection is reported to the queuing machine, that is, the sentiment is reported. / or the result of keyword detection to the queue.

508. The queuing machine notifies the CTI of the event message detected in real time.

509. The CTI issues the detected event message to the attendant and/or the quality inspector respectively.

Specifically, the CTI sends the detected event message to the operator and/or the quality inspector corresponding to the quality inspection task according to the setting of the quality inspection task; then the CTI business system can according to the event message. Carry out the corresponding business prompts, or the quality inspector performs the corresponding business processing such as the insert operation.

For a specific application example, as shown in FIG. 6 and FIG. 7 is a schematic diagram of an application using the method corresponding to the embodiment of the present invention in FIG. 5, and FIG. 6 is an interface for setting an emotion detection quality inspection task in advance; FIG. 7 is a After the emotion detection server detects in real time, the emotion detection event message displayed on the agent interface and the quality inspector's monitoring interface.

In this way, through the implementation of the technical solution of the above embodiment, real-time monitoring of the call content can be realized through the call center platform, thereby improving the service quality of the real-time call, and promoting the work improvement of the operator, thereby improving the service quality of the call center.

The embodiment of the present invention further provides a device for voice detection of a call center, and FIG. 8 is a schematic structural diagram of an apparatus according to an embodiment of the present invention. The device includes:

The call call recording starting unit 801 is configured to initiate a call call recording by the recording device 802 during the conversation between the operator and the user, and transmit the voice media stream generated for the attendant in real time to the voice detection server, so as to facilitate the The voice detection server generates an event message in real time according to the received voice media stream.

The recording device 802 is configured to start a call call recording according to the request of the call call recording starting unit 801, and generate a voice media stream for the attendant in real time.

The event message processing unit 803 is configured to receive an event message generated by the voice detection server, and issue a corresponding event prompt to the operator and/or the quality inspector.

In addition, the device may further include a quality inspection task setting unit 804. FIG. 9 is another schematic structural diagram of the device provided by the embodiment of the present invention, and FIG. 9 further includes:

The quality inspection task setting unit 804 is configured to preset a quality inspection task for voice detection. The call call recording activation unit 801 requests the recording device 802 according to the quality inspection task preset by the quality inspection task setting unit. Perform voice detection on the call.

The working process of each unit in the device is as described in the method corresponding to FIG. 5 above, and the real-time monitoring of the call content can be realized by implementing the technical solution of the device embodiment, thereby improving the service quality of the real-time call and promoting The operator's work has improved, which has improved the service quality of the call center. the amount.

Another embodiment of the present invention provides a device for voice detection of a call center. FIG. 10 is a schematic structural diagram of another device according to an embodiment of the present invention. The device includes:

The voice media stream receiving unit 1001 is configured to receive, during a call between the operator and the user, a voice media stream sent by the call center platform in real time for the attendant.

The event message generating unit 1002 is configured to detect an emotional point and/or a keyword in real time according to the received voice media stream, and generate a corresponding event message according to the detected emotional point and/or keyword.

The event message sending unit 1003 is configured to send the generated event message to the call center platform.

The working process of each unit in the device is as described in the method corresponding to FIG. 5 above, and the real-time monitoring of the call content can be realized by implementing the technical solution of the device embodiment, thereby improving the service quality of the real-time call and promoting The operator's work has improved, which in turn has improved the quality of service in the call center.

The embodiment of the present invention further provides a call center voice detection system. FIG. 11 is a schematic structural diagram of a system according to an embodiment of the present invention. The system includes a call center platform 1101 and a voice detection server 1 102, where:

The call center platform 1101 is configured to initiate a call call recording by the recording device during the call between the operator and the user, and transmit the voice media stream generated for the attendant in real time to the voice detection server 1102; The event message generated by the voice detection server 1102 is sent to the attendant and/or the quality inspector;

The voice detection server 1102 is configured to receive, during a call between the operator and the user, a voice media stream sent by the call center platform 1 101 for the attendant in real time; and then detect the emotional point and the real-time according to the received voice media stream. And/or keywords, and generating corresponding event messages according to the detected emotional points and/or keywords, and transmitting the generated event messages to the call center platform 1 101.

The working process of each part in the above system is as described in the method corresponding to FIG. 5 above, through the system By implementing the technical solution of the embodiment, real-time monitoring of the call content can be realized, thereby improving the service quality of the real-time call, and promoting the work improvement of the operator, thereby improving the service quality of the call center.

It should be noted that, in the foregoing apparatus and system embodiment, each unit included is only divided according to functional logic, but is not limited to the above division, as long as the corresponding function can be implemented; The specific names are also for convenience of distinguishing from each other and are not intended to limit the scope of the present invention.

In addition, those skilled in the art may understand that all or part of the steps in implementing the above method embodiments may be performed by a program to instruct related hardware, and the corresponding program may be stored in a computer readable storage medium, as mentioned above. The storage medium may be a read only memory, a magnetic disk or an optical disk or the like.

In summary, the embodiment of the present invention can realize real-time monitoring of the call content, thereby improving the service quality of the real-time call, and promoting the work improvement of the operator, thereby improving the service quality of the call center.

The above is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of it within the technical scope disclosed by the embodiments of the present invention. Variations or substitutions are intended to be covered by the scope of the invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

Claim

A method for voice detection in a call center, the method comprising:

During the conversation between the operator and the user, the call center platform initiates the call call recording through the recording device, and transmits the voice media stream generated for the attendant in real time to the voice detection server, so that the voice detection server receives the voice detection server according to the received voice call server. The voice media stream generates an event message in real time; the call center platform receives the event message generated by the voice detection server, and sends a corresponding event prompt to the operator and/or the quality inspector.

2. The method according to claim 1, wherein the call center platform initiates a call call recording by using a recording device, including:

The call center platform requests the recording device to perform voice detection on the call according to the preset quality inspection task, and the recording device starts the call call recording according to the request.

3. The method according to claim 2, wherein the preset quality inspection task comprises one or more of the following:

Specify the quality of the call, the quality of the specified time period, and the quality of the designated attendant.

4. The method of claim 1, wherein the recording device comprises:

Call center recording device, local agent recording device or IP capture recording device.

A method for voice detection in a call center, the method comprising:

Receiving a voice media stream for the attendant sent by the call center platform in real time during the conversation between the operator and the user;

The generated event message is sent to the call center platform.

The method according to claim 5, wherein the generating the corresponding event message according to the detected emotional point and/or keyword comprises:

If a preset emotional point and/or keyword is detected, an event message for the emotional point and/or keyword is generated; If no pre-set mood points and/or keywords are detected, no event message is generated or an event message with a null result is generated.

7. A device for voice detection in a call center, comprising:

a recording device, configured to start a call call recording according to the request of the call call recording starting unit, and generate a voice media stream for the attendant in real time;

The event message processing unit is configured to receive an event message generated by the voice detection server, and send a corresponding event prompt to the operator and/or the quality inspector.

8. The device according to claim 7, wherein the device further comprises:

The quality inspection task setting unit is configured to preset a quality inspection task of voice detection;

The call call recording activation unit requests the recording device to perform voice detection on the call according to the quality inspection task preset by the quality inspection task setting unit.

9. A device for voice detection in a call center, comprising:

An event message sending unit is configured to send the generated event message to the call center platform.

A call center voice detection system, wherein the system comprises a call center platform and a voice detection server, wherein:

The call center platform is configured to initiate a call call recording by the recording device during the call between the operator and the user, and transmit the voice media stream generated for the attendant in real time to the voice detection server; and receive the voice detection The event message generated by the server, and the corresponding thing is issued a reminder to the operator and / or quality inspector;