CN111835925A - Off-line voice quality inspection and analysis system for call center - Google Patents
Off-line voice quality inspection and analysis system for call center Download PDFInfo
- Publication number
- CN111835925A CN111835925A CN202010548281.XA CN202010548281A CN111835925A CN 111835925 A CN111835925 A CN 111835925A CN 202010548281 A CN202010548281 A CN 202010548281A CN 111835925 A CN111835925 A CN 111835925A
- Authority
- CN
- China
- Prior art keywords
- quality inspection
- data
- call
- module
- voice
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04M—TELEPHONIC COMMUNICATION
- H04M3/00—Automatic or semi-automatic exchanges
- H04M3/42—Systems providing special services or facilities to subscribers
- H04M3/50—Centralised arrangements for answering calls; Centralised arrangements for recording messages for absent or busy subscribers ; Centralised arrangements for recording messages
- H04M3/51—Centralised call answering arrangements requiring operator intervention, e.g. call or contact centers for telemarketing
- H04M3/5175—Call or contact centers supervision arrangements
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
- G10L25/63—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination for estimating an emotional state
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Child & Adolescent Psychology (AREA)
- Hospice & Palliative Care (AREA)
- Psychiatry (AREA)
- Business, Economics & Management (AREA)
- Marketing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Telephonic Communication Services (AREA)
Abstract
The invention provides an off-line voice quality inspection and analysis system facing a call center, which comprises: the system comprises a data acquisition module, a data preprocessing module, a voice conversion module, an intelligent quality inspection module, an intelligent analysis module, a Web application system module and a data storage module, wherein other modules interact with the data storage module to read and write data. The voice quality inspection method can improve the quality inspection coverage rate to 100%, greatly improve the voice quality inspection efficiency, solve the limitation of manual quality inspection, improve the service quality and management level, reduce the operation cost of enterprises, and assist business operation decision.
Description
Technical Field
The invention relates to the technical field of voice quality inspection, in particular to an offline voice quality inspection and analysis system for a call center.
Background
Call centers have a large amount of call audio data every day, and agent personnel handle various needs of customers, such as: after sale, complaints, consultations and the like, the high-quality seat service can not only improve the enterprise image but also reduce the customer loss rate. In order to improve the service quality of the seat, a team leader or a supervisor can judge whether the words of the seat meet the service standard or not by means of spot check and listening to the call recording file. The quality inspection of the existing call center is more manual quality inspection. The call center stores the call audio as a recording file, quality testing personnel extract 1-5% of the recording file to perform tuning detection, detect whether the initial words and the final words are standard or not, detect whether the customer service personnel have dirty words or not and feed back the words to the personnel in the seat, and improve the service quality of the personnel in the seat. In addition, the data can be simply arranged and summarized only by means of a table or the like. The existing quality inspection technology has the following problems:
quality testing personnel carry out quality testing according to the understanding of the quality testing standard document, and the quality testing result is greatly influenced by subjective factors and lacks of objectivity.
The department leader monitors and patrols to find problems or the customer service staff actively seeks help, most problems cannot be found and solved in time, and timeliness is poor.
After-the-fact spot check, the spot check amount accounts for 1 to 5 percent of the total amount, and the coverage rate of quality check is low; the proportion of quality inspectors is 1:40, each person can inspect dozens of quality inspectors at most every day, and the quality inspection efficiency is low.
The method can only be used for arranging and summarizing simple data by means of tools such as tables and the like, cannot fully mine mass recording data, and has low data utilization value.
Disclosure of Invention
Aiming at the problems of large human factor interference and lack of objectivity existing in the manual spot check and identification of the seat words in the prior art, the invention provides an offline voice quality inspection and analysis system for a call center, which can improve the coverage rate of quality inspection to 100%, greatly improve the voice quality inspection efficiency, solve the limitation of manual quality inspection, improve the service quality and management level, reduce the operation cost of enterprises and assist business operation decision making.
In order to achieve the above purpose, the present invention provides the following technical solutions:
an off-line voice quality inspection and analysis system for a call center comprises: the data acquisition module acquires call recording data and corresponding channel associated data from the call center, stores the acquired data in the data storage module and generates a first call identifier; the data preprocessing module is used for receiving the first call identifier, acquiring call recording data and channel associated data from the data storage module for preprocessing to obtain preprocessed data, storing the preprocessed data into the data storage module and generating a second call identifier; the voice transcription module is used for receiving the second communication identifier, acquiring the preprocessed data from the data storage module, performing voice transcription, speaker separation, role distinguishing and paragraph processing on the preprocessed data to obtain transcribed data, storing the transcribed data in the data storage module and generating a third communication identifier; the intelligent quality inspection module receives the third communication identifier, acquires the transcribed data from the data storage module, creates a quality inspection template, performs quality inspection on the transcribed data, generates a quality inspection result and stores the quality inspection result in the data storage module; the intelligent analysis module receives the third call identifier, acquires the transcribed data from the data storage module, performs semantic analysis on the transcribed data, judges the reason of the incoming call of the user and the service type of the incoming call, obtains an analysis result, and stores the quality inspection result in the data storage module; the Web application system module is used for performing man-machine interaction with a user, acquiring an analysis result and a quality inspection result from the data storage module and displaying the analysis result and the quality inspection result in a visual mode; and the data storage module is used for storing and calling data of each module.
The data acquisition module and the voice transcription module realize the structured processing of unstructured call audio data, can perform indexing call retrieval, and the data preprocessing module provides audio format transcription and voice enhancement, ensures the accuracy of ASR voice transcription, and the intelligent analysis module performs text classification, abnormal emotion detection, named object recognition and hot word analysis by using machine learning and NLP related algorithms, such as Bert, HiGRU, BilSTM-CRF, TextRank and the like, deeply excavates call potential information and provides data support for enterprise operation strategies; a Web application system module: the quality inspection system is visual, a user can retrieve call data on a front-end page, set quality inspection rules, inquire quality inspection results, manually review, mention complaints, process complaints and the like, and the module also displays quality inspection statistical information based on intelligent quality inspection and operation statistical information based on intelligent service analysis. Full quality inspection, the system ensures full quality inspection of the recorded data of the call center; the quality inspection is fair, the intelligent quality inspection ensures the fairness of the quality inspection, and the quality inspection result is not influenced by subjective factors;
preferably, the transcription data includes audio attributes, associated data, transcription text, character information, and paragraph information.
Preferably, the preprocessing includes transcoding the recording data, adjusting a sampling rate, adjusting a volume, reducing noise, and performing format conversion, unit conversion, and service processing on the associated data.
Preferably, the intelligent quality inspection module comprises a keyword quality inspection unit, based on a keyword library set in a quality inspection template by a user, whether the keyword library comprises a precondition, a quality inspection range and whether certain keywords, the system detects a call transcribed text, and if the keywords are matched in a fuzzy manner, the transcribed text is brought into a quality inspection result; and the speech speed quality inspection unit detects the speech speed of each sentence spoken by the seat personnel in the call based on a speech speed threshold value set in the quality inspection template by the user, and the speech speed exceeds the threshold value to indicate hit and is included in a quality inspection result.
Preferably, the intelligent quality inspection module comprises a preemptive call quality inspection unit, and the intelligent quality inspection module detects whether the agent carries out the preemptive call when the client speaks or indicates hit when the preemptive call exceeds the threshold value based on a preemptive call cross word number threshold value set in the quality inspection template by the user, and incorporates a quality inspection report; and the mute quality inspection unit is used for detecting whether the agent timely responds to the client and whether the agent is in a cold field in the call service process or not based on a mute time threshold set in the quality inspection template by the user, and the mute time exceeds the threshold to indicate hit and bring the mute time into a quality inspection result.
Preferably, the intelligent quality inspection module comprises a call duration quality inspection unit, and detects the duration of a call based on a call duration threshold set in the quality inspection template by a user, wherein the duration exceeds the threshold, namely the call duration is hit, and a quality inspection result is included; and the abnormal emotion quality detection unit detects the emotion of the corresponding role in the call information based on the abnormal emotion detection object set in the quality detection template by the user, predicts negative emotion, namely, shows hit, and brings the abnormal emotion into a quality detection result.
Preferably, the intelligent quality inspection module comprises a service aversion quality inspection unit, the conversation transliteration text is detected based on a service aversion library set in the quality inspection template by the user, if the service aversion is matched, the conversation transliteration text is hit, and the hit service aversion is included in the quality inspection result.
Preferably, the intelligent analysis module comprises an incoming call analysis unit, and analyzes the reason of repeated incoming calls and brings the reason into an analysis result by counting incoming calls of the same number for multiple times in a certain time period and consulting the same service; and the hot word adjusting unit is used for making hot word statistics and analyzing results by counting high-frequency words and high-weight words appearing in a period of time. The statistical analysis is comprehensive, and comprises quality inspection overview, grading distribution, quality inspection rules, call duration analysis, mute analysis, repeated incoming calls, incoming call reasons, service types, hot word analysis and the like.
Preferably, the intelligent quality inspection module comprises an alarm unit, and based on an alarm range set in the quality inspection template by a user, the intelligent quality inspection module sends alarm information in time when the preset alarm range is hit; the alarm mode comprises the following steps: short message alarm, nail alarm, and third-party system alarm.
A large amount of recorded call data are intelligently and automatically converted into texts through voice, keyword detection, speech speed analysis, mute analysis, emotion detection and the like, potential information of the transcribed texts is mined through an intelligent analysis system, and data support is provided for enterprise operation decisions. And the data is visually displayed, and the quality inspection overview of the seat personnel, the service analysis and the like can be checked in the web system.
The invention has the following beneficial effects: full quality inspection, the system ensures full quality inspection of the recorded data of the call center; the quality inspection is fair, the intelligent quality inspection ensures the fairness of the quality inspection, and the quality inspection result is not influenced by subjective factors; and (3) deep intelligence, namely applying the semantic understanding NLU to voice quality inspection, and performing text classification, abnormal emotion detection, named body recognition and hotword analysis by using Bert, HiGRU, BilSTM-CRF, TextRank and the like. The quality inspection based on semantic understanding is more advanced than the quality inspection based on rules, and the quality inspection result is more accurate. The statistical analysis is comprehensive, and comprises quality inspection overview, grading distribution, quality inspection rules, call duration analysis, mute analysis, repeated incoming calls, incoming call reasons, service types, hot word analysis and the like.
Drawings
Fig. 1 is a system configuration diagram of the present embodiment.
Detailed Description
The present embodiment provides an off-line voice quality inspection and analysis system for a call center, referring to fig. 1, including: the data acquisition module acquires call recording data and corresponding channel associated data from the call center, stores the acquired data in the data storage module and generates a first call identifier; the data preprocessing module is used for receiving the first call identifier, acquiring call recording data and channel associated data from the data storage module for preprocessing, wherein the preprocessing comprises transcoding the recording data, adjusting the sampling rate, adjusting the volume, reducing noise and the like, performing format conversion, unit conversion and service processing on the channel associated data to obtain preprocessed data, storing the preprocessed data in the data storage module and generating a second call identifier; the voice transcription module is used for receiving the second communication identifier, acquiring the preprocessed data from the data storage module, performing voice transcription, speaker separation, role distinguishing and paragraph processing on the preprocessed data to obtain transcribed data, storing the transcribed data in the data storage module and generating a third communication identifier; the transcription data includes audio attributes, channel associated data, transcription text, character information, and paragraph information.
The intelligent quality inspection module receives the third communication identifier, acquires the transcribed data from the data storage module, creates a quality inspection template, performs quality inspection on the transcribed data, generates a quality inspection result and stores the quality inspection result in the data storage module; the intelligent analysis module receives the third call identifier, acquires the transcribed data from the data storage module, performs semantic analysis on the transcribed data, judges the reason of the incoming call of the user and the service type of the incoming call, obtains an analysis result, and stores the quality inspection result in the data storage module; the Web application system module is used for performing man-machine interaction with a user, acquiring an analysis result and a quality inspection result from the data storage module and displaying the analysis result and the quality inspection result in a visual mode; and the data storage module is used for storing and calling data of each module.
The data acquisition module and the voice transcription module realize the structured processing of unstructured call audio data, can perform indexing call retrieval, and the data preprocessing module provides audio format transcription and voice enhancement, ensures the accuracy of ASR voice transcription, and the intelligent analysis module performs text classification, abnormal emotion detection, named object recognition and hot word analysis by using machine learning and NLP related algorithms, such as Bert, HiGRU, BilSTM-CRF, TextRank and the like, deeply excavates call potential information and provides data support for enterprise operation strategies; a Web application system module: the quality inspection system is visual, a user can retrieve call data on a front-end page, set quality inspection rules, inquire quality inspection results, manually review, mention complaints, process complaints and the like, and the module also displays quality inspection statistical information based on intelligent quality inspection and operation statistical information based on intelligent service analysis. Full quality inspection, the system ensures full quality inspection of the recorded data of the call center; the quality inspection is fair, the intelligent quality inspection ensures the fairness of the quality inspection, and the quality inspection result is not influenced by subjective factors;
the intelligent quality inspection module comprises a keyword quality inspection unit, a speech speed quality inspection unit, a call-robbing and call-inserting quality inspection unit, a mute quality inspection unit, a call duration quality inspection unit, an abnormal emotion quality inspection unit, a service avoiding speech quality inspection unit and an alarm unit; the intelligent analysis module comprises an incoming call analysis unit.
The keyword quality inspection unit is used for detecting the conversation transcribed text by the system based on a keyword library set in the quality inspection template by the user and including whether the keyword library contains preconditions, quality inspection ranges and whether certain keywords or not, and if the keywords are matched in a fuzzy manner, the transcribed text is brought into a quality inspection result;
and the speech speed quality inspection unit detects the speech speed of each sentence spoken by the seat personnel in the call based on a speech speed threshold value set in the quality inspection template by the user, and the speech speed exceeds the threshold value to indicate hit and is included in a quality inspection result.
The quality inspection unit for the preemptive interposing words detects whether the agent carries out interposing when the client speaks or preemptively calls after the client speaks, and the preemptive interposing words exceed the threshold value to indicate hitting and bring in a quality inspection report based on the preemptive interposing words threshold value set in the quality inspection template by the user;
and the mute quality inspection unit is used for detecting whether the agent timely responds to the client and whether the agent is in a cold field in the call service process or not based on a mute time threshold set in the quality inspection template by the user, and the mute time exceeds the threshold to indicate hit and bring the mute time into a quality inspection result.
The call duration quality inspection unit is used for detecting the duration of one call based on a call duration threshold value set in the quality inspection template by a user, wherein the duration exceeds the threshold value, namely, the call duration is hit, and a quality inspection result is included;
and the abnormal emotion quality detection unit detects the emotion of the corresponding role in the call information based on the abnormal emotion detection object set in the quality detection template by the user, predicts negative emotion, namely, shows hit, and brings the abnormal emotion into a quality detection result.
And the service aversion word quality detection unit is used for detecting the call transcribed text based on a service aversion word library set in the quality detection template by the user, indicating a hit if the service aversion word is matched, and bringing the hit service aversion word into a quality detection result.
The warning unit is used for sending warning information in time when a preset warning range is hit based on the warning range set in the quality inspection template by the user; the alarm mode comprises the following steps: short message alarm, nail alarm, and third-party system alarm.
The incoming call analysis unit analyzes the reason of repeated incoming calls and brings the analysis result into consideration by counting multiple incoming calls of the same number for a certain period of time and consulting the same service; and the hot word adjusting unit is used for making hot word statistics and analyzing results by counting high-frequency words and high-weight words appearing in a period of time. The statistical analysis is comprehensive, and comprises quality inspection overview, grading distribution, quality inspection rules, call duration analysis, mute analysis, repeated incoming calls, incoming call reasons, service types, hot word analysis and the like.
The embodiment further provides an offline voice quality inspection and analysis method facing the call center, and the offline voice quality inspection and analysis system facing the call center specifically includes the following steps:
s1, obtaining the call recording data and the corresponding associated data from the call center;
s2, preprocessing the call recording data and the channel associated data to obtain preprocessed data;
the preprocessing comprises transcoding the recorded data, adjusting the sampling rate, adjusting the volume, reducing noise and the like, and performing format conversion, unit conversion and service processing on the channel associated data;
s3, performing voice transcription, speaker separation, role differentiation and paragraph processing on the preprocessed data to obtain transcribed data, wherein the transcribed data comprises audio attributes, channel associated data, transcribed texts, role information and paragraph information;
s4, the created quality inspection template performs quality inspection on the data to be transferred and written to generate a quality inspection result;
s5, performing semantic analysis on the data to be transferred, judging the reason of the incoming call of the user and the service type of the incoming call, and obtaining an analysis result; by counting a certain time period, the same number of incoming calls consults the same service for multiple times, analyzes the reason of repeated incoming calls and brings the reason into the analysis result; and performing hot word statistics and analyzing results by counting high-frequency words and high-weight words appearing in a period of time.
Step S4 specifically includes:
s401, performing keyword quality inspection, wherein the system detects a call transcribed text based on a keyword library set in a quality inspection template by a user and including whether the pre-condition, the quality inspection range and certain keywords are included, and if the keywords are matched in a fuzzy manner, the transcribed text is included in a quality inspection result;
s402, speech rate quality inspection, based on the speech rate threshold set in the quality inspection template by the user, detecting the speech rate of each sentence spoken by the agent in the call, and if the speech rate exceeds the threshold, indicating hit, and bringing the speech rate into the quality inspection result.
S403, performing urgent insertion speech quality inspection, detecting whether the agent performs the speech insertion when the client speaks or not based on an urgent insertion speech cross word number threshold value set in a quality inspection template by the user, or detecting whether the speech insertion is performed when the speech number exceeds the threshold value after the speech insertion of the client, and bringing a quality inspection report into the inspection result;
s404, performing mute quality inspection, namely detecting whether the agent timely responds to the client and whether the agent is in a cold field in the call service process or not based on a mute time threshold set in a quality inspection template by the user, wherein the condition that the mute time exceeds the threshold indicates hit and the quality inspection result is included.
S405, call duration quality inspection, wherein the call duration of one call is detected based on a call duration threshold set in a quality inspection template by a user, and the call duration exceeds the threshold, namely hit is indicated, and a quality inspection result is included;
s406, abnormal emotion quality inspection, wherein the emotion of the corresponding role in the call information is detected based on an abnormal emotion detection object set in a quality inspection template by the user, the abnormal emotion detection unit predicts negative emotion, namely, hit, and brings the abnormal emotion into a quality inspection result.
And S407, performing quality inspection on the service aversion, detecting a call transliteration text based on a service aversion library set in a quality inspection template by a user, and if the service aversion is matched, indicating a hit, and incorporating the hit service aversion into a quality inspection result.
Compared with manual spot check, the method has the advantages of achieving full quality check, fair quality check, automatic quality check and comprehensive statistics, improving the coverage rate of the quality check to 100%, greatly improving the voice quality check efficiency, solving the limitation of the manual quality check, improving the service quality and the management level, reducing the operation cost of enterprises and assisting business operation decision.
The invention has the following advantages: full quality inspection, the system ensures full quality inspection of the recorded data of the call center; the quality inspection is fair, the intelligent quality inspection ensures the fairness of the quality inspection, and the quality inspection result is not influenced by subjective factors; and (3) deep intelligence, namely applying the semantic understanding NLU to voice quality inspection, and performing text classification, abnormal emotion detection, named body recognition and hotword analysis by using Bert, HiGRU, BilSTM-CRF, TextRank and the like. The quality inspection based on semantic understanding is more advanced than the quality inspection based on rules, and the quality inspection result is more accurate. The statistical analysis is comprehensive, and comprises quality inspection overview, grading distribution, quality inspection rules, call duration analysis, mute analysis, repeated incoming calls, incoming call reasons, service types, hot word analysis and the like.
Claims (9)
1. An off-line voice quality inspection and analysis system for a call center is characterized by comprising:
the data acquisition module acquires call recording data and corresponding channel associated data from the call center, stores the acquired data in the data storage module and generates a first call identifier;
the data preprocessing module is used for receiving the first call identifier, acquiring call recording data and channel associated data from the data storage module for preprocessing to obtain preprocessed data, storing the preprocessed data into the data storage module and generating a second call identifier;
the voice transcription module is used for receiving the second communication identifier, acquiring the preprocessed data from the data storage module, performing voice transcription, speaker separation, role distinguishing and paragraph processing on the preprocessed data to obtain transcribed data, storing the transcribed data in the data storage module and generating a third communication identifier;
the intelligent quality inspection module receives the third communication identifier, acquires the transcribed data from the data storage module, creates a quality inspection template, performs quality inspection on the transcribed data, generates a quality inspection result and stores the quality inspection result in the data storage module;
the intelligent analysis module receives the third call identifier, acquires the transcribed data from the data storage module, performs semantic analysis on the transcribed data, judges the reason of the incoming call of the user and the service type of the incoming call, obtains an analysis result, and stores the quality inspection result in the data storage module;
the Web application system module is used for performing man-machine interaction with a user, acquiring an analysis result and a quality inspection result from the data storage module and displaying the analysis result and the quality inspection result in a visual mode;
and the data storage module is used for storing and calling data of each module.
2. The system of claim 1, wherein the transcription data comprises audio attributes, channel associated data, transcription text, role information and paragraph information.
3. The system of claim 1, wherein the preprocessing comprises transcoding the recorded data, adjusting sampling rate, adjusting volume, reducing noise, converting format, converting unit, and processing service of the associated data.
4. The system of claim 1, wherein the intelligent quality inspection module comprises a voice module for testing and analyzing the voice of the call center
The keyword quality inspection unit is used for detecting the conversation transcribed text by the system based on a keyword library set in the quality inspection template by the user and including whether the keyword library contains preconditions, quality inspection ranges and whether certain keywords or not, and if the keywords are matched in a fuzzy manner, the transcribed text is brought into a quality inspection result;
and the speech speed quality inspection unit detects the speech speed of each sentence spoken by the seat personnel in the call based on a speech speed threshold value set in the quality inspection template by the user, and the speech speed exceeds the threshold value to indicate hit and is included in a quality inspection result.
5. The system of claim 1, wherein the intelligent quality inspection module comprises a voice module for testing and analyzing the voice of the call center
The quality inspection unit for the preemptive interposing words detects whether the agent carries out interposing when the client speaks or preemptively calls after the client speaks, and the preemptive interposing words exceed the threshold value to indicate hitting and bring in a quality inspection report based on the preemptive interposing words threshold value set in the quality inspection template by the user;
and the mute quality inspection unit is used for detecting whether the agent timely responds to the client and whether the agent is in a cold field in the call service process or not based on a mute time threshold set in the quality inspection template by the user, and the mute time exceeds the threshold to indicate hit and bring the mute time into a quality inspection result.
6. The system of claim 1, wherein the intelligent quality inspection module comprises a voice module for testing and analyzing the voice of the call center
The call duration quality inspection unit is used for detecting the duration of one call based on a call duration threshold value set in the quality inspection template by a user, wherein the duration exceeds the threshold value, namely, the call duration is hit, and a quality inspection result is included;
and the abnormal emotion quality detection unit detects the emotion of the corresponding role in the call information based on the abnormal emotion detection object set in the quality detection template by the user, predicts negative emotion, namely, shows hit, and brings the abnormal emotion into a quality detection result.
7. The offline voice quality inspection and analysis system facing the call center as claimed in claim 1, wherein the intelligent quality inspection module comprises a service aversion quality inspection unit, detects a conversation transliteration text based on a service aversion library set in a quality inspection template by a user, and if the service aversion is matched, the conversation transliteration text is hit, and the hit service aversion is included in a quality inspection result.
8. The system of claim 1, wherein the intelligent analysis module comprises a voice module for performing voice quality inspection and analysis, and the voice module comprises a voice module for performing voice quality inspection and analysis
The incoming call analysis unit analyzes the reason of repeated incoming calls and brings the analysis result into consideration by counting multiple incoming calls of the same number for a certain period of time and consulting the same service;
and the hot word adjusting unit is used for making hot word statistics and analyzing results by counting high-frequency words and high-weight words appearing in a period of time.
9. The off-line voice quality inspection and analysis system for the call center as claimed in claim 1, wherein the intelligent quality inspection module comprises an alarm unit for sending an alarm message in time when a preset alarm range is hit based on the alarm range set in the quality inspection template by a user; the alarm mode comprises the following steps: short message alarm, nail alarm, and third-party system alarm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010548281.XA CN111835925A (en) | 2020-06-16 | 2020-06-16 | Off-line voice quality inspection and analysis system for call center |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010548281.XA CN111835925A (en) | 2020-06-16 | 2020-06-16 | Off-line voice quality inspection and analysis system for call center |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111835925A true CN111835925A (en) | 2020-10-27 |
Family
ID=72898675
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010548281.XA Pending CN111835925A (en) | 2020-06-16 | 2020-06-16 | Off-line voice quality inspection and analysis system for call center |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111835925A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112511698A (en) * | 2020-12-03 | 2021-03-16 | 普强时代(珠海横琴)信息技术有限公司 | Real-time call analysis method based on universal boundary detection |
| CN113723975A (en) * | 2021-09-13 | 2021-11-30 | 国泰君安证券股份有限公司 | System, method, device, processor and computer readable storage medium for realizing intelligent quality inspection processing in intelligent return visit service |
| CN114842849A (en) * | 2022-04-24 | 2022-08-02 | 马上消费金融股份有限公司 | Voice conversation detection method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106021532A (en) * | 2016-05-25 | 2016-10-12 | 东软集团股份有限公司 | Display method and device for keywords |
| CN108388558A (en) * | 2018-02-07 | 2018-08-10 | 平安普惠企业管理有限公司 | Problem matching process, device, customer service robot and storage medium |
| CN109145281A (en) * | 2017-06-15 | 2019-01-04 | 北京嘀嘀无限科技发展有限公司 | Audio recognition method, device and storage medium |
-
2020
- 2020-06-16 CN CN202010548281.XA patent/CN111835925A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106021532A (en) * | 2016-05-25 | 2016-10-12 | 东软集团股份有限公司 | Display method and device for keywords |
| CN109145281A (en) * | 2017-06-15 | 2019-01-04 | 北京嘀嘀无限科技发展有限公司 | Audio recognition method, device and storage medium |
| CN108388558A (en) * | 2018-02-07 | 2018-08-10 | 平安普惠企业管理有限公司 | Problem matching process, device, customer service robot and storage medium |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112511698A (en) * | 2020-12-03 | 2021-03-16 | 普强时代(珠海横琴)信息技术有限公司 | Real-time call analysis method based on universal boundary detection |
| CN113723975A (en) * | 2021-09-13 | 2021-11-30 | 国泰君安证券股份有限公司 | System, method, device, processor and computer readable storage medium for realizing intelligent quality inspection processing in intelligent return visit service |
| CN114842849A (en) * | 2022-04-24 | 2022-08-02 | 马上消费金融股份有限公司 | Voice conversation detection method and device |
| CN114842849B (en) * | 2022-04-24 | 2023-08-08 | 马上消费金融股份有限公司 | Voice dialogue detection method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7596498B2 (en) | Monitoring, mining, and classifying electronically recordable conversations | |
| CN111835925A (en) | Off-line voice quality inspection and analysis system for call center | |
| US7599475B2 (en) | Method and apparatus for generic analytics | |
| US8050923B2 (en) | Automated utterance search | |
| JP4662861B2 (en) | Monitoring device, evaluation data selection device, respondent evaluation device, respondent evaluation system and program | |
| US8676586B2 (en) | Method and apparatus for interaction or discourse analytics | |
| US8150021B2 (en) | Proactive system and method for monitoring and guidance of call center agent | |
| US9105042B2 (en) | Customer sentiment analysis using recorded conversation | |
| CN109448730A (en) | A kind of automatic speech quality detecting method, system, device and storage medium | |
| US20100246799A1 (en) | Methods and apparatus for deep interaction analysis | |
| CN105975514A (en) | Automatic quality testing method and system | |
| CN113468296B (en) | Model self-iteration type intelligent customer service quality inspection system and method capable of configuring business logic | |
| CA2567230A1 (en) | Multi-pass speech analytics | |
| JP2011082839A (en) | Call center system | |
| CN110135669A (en) | A kind of service quality inspection management method and system | |
| CN104301554A (en) | Device and method used for detecting service quality of customer service staff | |
| CN111294468A (en) | Tone quality detection and analysis system for customer service center calling | |
| CN110890096A (en) | Intelligent voice system and method based on voice analysis | |
| CN111951831A (en) | Method for realizing audio quality inspection based on AI | |
| CN111128241A (en) | Intelligent quality inspection method and system for voice call | |
| CN112258016A (en) | Customer service intelligent quality inspection analysis method based on ASR | |
| CN113361969B (en) | Intelligent quality inspection system capable of flexibly configuring templates | |
| CN111147669A (en) | Full real-time automatic service quality inspection system and method | |
| CN112261230B (en) | Express call answering rate assessment method, device, equipment, system and storage medium | |
| CN113011159A (en) | Artificial seat monitoring method and device, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20201027 |