Character transfer and quality detection analysis system based on telephone recording

By developing a quality detection and analysis system based on voice recognition and language model in the call center system, the problem that traditional artificial quality inspection methods are difficult to meet large-scale data analysis is solved, and automated and intelligent call recording quality inspection is realized, which improves efficiency and accuracy, improving customer satisfaction and service quality.

CN118945269AInactive Publication Date: 2024-11-12POINT CONTROL CLOUD (BEIJING) INTELLIGENT TECH CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202410806010.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-06-21
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Traditional artificial quality inspection methods are difficult to cover all recorded data, cannot meet the quality evaluation and monitoring needs of large-scale call data, and are time-consuming and labor-intensive, and lack automated and intelligent quality inspection capabilities.

Method used

Develop a text transcription and quality detection and analysis system based on telephone recording, using advanced speech recognition algorithms, noise reduction algorithms, language model technology and visual reports to realize automated full-scale call recording data quality detection and analysis, and support flexible definition of quality Detect templates, capture intonation and emotional changes, and generate quality detection results.

Benefits of technology

It improves the efficiency, accuracy and consistency of quality inspection work, reduces costs, ensures that service quality meets standards, improves customer satisfaction, and provides accurate customer sentiment analysis support.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure FT_1
    Figure FT_1
Patent Text Reader

Abstract

The invention discloses a character transfer and quality detection and analysis system based on telephone recording, and relates to an internet technology, a voice conversion technology and a contact communication system, in particular to a quality detection and analysis method. The core object of the invention is to solve the practical problems mentioned in the background. Through the innovative method and technology, the data analysis quality is improved, the quality inspection work efficiency, accuracy and consistency are improved, the cost is reduced, and it is ensured that relevant standards and laws and regulations are met. The external service quality of the call center is improved, and the customer satisfaction is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to Internet technology, voice conversion technology and contact communication system, and in particular to a quality detection and analysis method. Background Art

[0002] In my country, call center systems have a large and widespread user base. As the number of calls made by agents through call center systems increases, companies are faced with more and more call data that needs to be evaluated and monitored for quality. Traditional manual quality inspection methods are often time-consuming and labor-intensive, and it is difficult to cover all the recording data, and they cannot meet the needs of large-scale data analysis.

[0003] To solve this problem, the inventor proposed a development plan for the invention, a text transcription and quality detection and analysis system based on telephone recordings. It can realize automatic and intelligent quality detection and analysis of the entire call recording data, filling the gap of traditional manual quality inspection, and can automatically evaluate the quality of massive data, improving work efficiency and accuracy. Summary of the invention

[0004] The core goal of the present invention is to solve the practical problems mentioned in the above background. Through innovative methods and technologies, the quality of data analysis is improved, the efficiency, accuracy and consistency of quality inspection work are improved, costs are reduced, and compliance with relevant standards and regulations is ensured. The quality of external services of the call center is improved and customer satisfaction is increased.

[0005] Compared with the existing methods, the beneficial effects of the present invention are:

[0006] 1. High accuracy: The system integrates advanced speech recognition algorithms and noise reduction algorithms to enable it to accurately identify the content and speech speed of recordings, ensuring highly accurate text transcription even in complex environments, and improving the overall quality inspection benchmark effect.

[0007] 2. Result analysis: After completing the text transcription, the system can quickly analyze the conversation content, generate call quality detection results, and provide instant feedback to administrators to adjust communication strategies, thereby improving customer satisfaction and problem-solving efficiency; using language model technology, the system can capture the tone and emotional changes of both parties on the call, providing more accurate data support for companies to understand and analyze customer emotions.

[0008] 3. High flexibility: Enterprises can flexibly define quality inspection templates based on industry characteristics and business needs to ensure that the system better adapts to the personalized demands of the enterprise and meets business expectations.

[0009] 4. Visual reports: Visual reports present data through charts, graphs and other intuitive methods, making complex data information easy to understand. Visual reports display analysis results and trends, allowing administrators to quickly understand information, better manage and optimize service processes, and provide data support for decision-making. DETAILED DESCRIPTION

[0010] Step 1:

[0011] Configure one or more quality detection templates based on the business type, call quality detection requirements, and compliance requirements of the enterprise call. The template supports configuring detection points and their corresponding scores, as well as matching quality detection tags through detection points and triggering quality detection warnings.

[0012] The score of the detection point needs to be set to a positive integer greater than or equal to 0 and less than or equal to 100, and the sum of the scores of all detection points is equal to 100.

[0013] The detection point configuration content includes: keyword detection, key word detection, regular expression matching detection, speaking speed detection, silence detection, interruption detection, emotion detection, context detection, trigger quality detection label, and trigger quality detection warning.

[0014] Step 2:

[0015] Connect to the call center system and configure the interface address information of the quality inspection and analysis system to obtain recordings, call records and other accompanying data from the call center system; configure the interface address information of the call center system to receive the quality inspection related result data returned by the quality inspection and analysis system

[0016] Step 3:

[0017] Filter recordings and initiate quality inspection tasks. The administrator can select the range of recordings for which he wants to initiate quality inspection tasks based on different filter items on the quality inspection and analysis system page.

[0018] The filter items include: the name of the connected call center system or company, the time period when the recording was made (accurate to the minute), the call type, and the recording duration.

[0019] After filling in the filter items, all recording items that meet the filter criteria will be displayed on the page. Administrators can check or select all recordings according to their needs, select the corresponding quality inspection template, and initiate quality inspection tasks.

[0020] Step 4:

[0021] After completing step 3, the quality inspection and analysis system will obtain basic recording and call record data based on the interface address information configured in step 2. After the data enters the system, automatic speech recognition is first performed to transcribe the audio content into text. After that, the system will use the transcribed text content to perform an intelligent comparison with the configuration in the quality inspection template selected in step 3 to generate quality inspection results.

[0022] The completion time of this step will fluctuate depending on the number of recordings and the total duration of recordings in the quality inspection task.

[0023] Step 5:

[0024] After step 4 is completed, the administrator can view, analyze, and modify the quality inspection results of the recordings in the management interface of the quality inspection and analysis system.

[0025] The quality inspection results include: the total quality inspection score, the score of each inspection point, the triggered quality inspection label, the triggered quality inspection warning, and the transcribed text content.

[0026] The administrator can choose in the management interface whether to return the quality inspection results to the call center system through the interface.

[0027] Step 6:

[0028] With the permission of the administrator, agents can view quality inspection results and file complaints in the call center system.

[0029] The administrator can choose to approve or reject the appeal in the quality inspection and analysis system. For approved appeal requests, the administrator will adjust the corresponding content and score in the quality inspection results and update the quality inspection results.

[0030] Step 7:

[0031] The quality inspection and analysis data and basic traffic data (from the call center system) generated by the quality inspection and analysis system will be analyzed visually in the statistics and analysis module of the quality inspection and analysis system through multiple functions such as overall report, agent report, inspection point report, keyword report, label analysis, and early warning analysis, so as to provide data support for the administrator's subsequent operations and decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is the data flow and analysis flow chart of this device.

Claims

1. A text transcription and quality detection and analysis system based on telephone recording, characterized in that: The system comprises: a template configuration unit, a data interface unit, a task initiation unit, a quality detection and analysis unit, a management interface unit, a complaint processing unit, and a data statistics and analysis unit.

2. The system according to claim 1, characterized in that The template configuration unit includes a template editor for customizing the score settings of the detection points, and the configuration of the associated quality detection labels and warning trigger conditions.

3. The system according to claim 1, characterized in that The data interface unit includes a data synchronization mechanism for real-time synchronization of recording data and quality detection result data of the call center system.

4. The system according to claim 1, characterized in that The task initiating unit includes a task scheduler, which is used to automatically initiate a quality inspection task based on the screening conditions of the administrator.

5. The system according to claim 1, characterized in that The quality detection and analysis unit includes a speech recognition processor and a quality detection comparator, which are used to transcribe the recordings and compare them with the quality detection templates to generate detection results.

6. The system according to claim 1, characterized in that The management interface unit includes a result feedback device and a user authority management mechanism. The result feedback device is used to display the quality inspection results to the administrator in the form of charts, graphics or text. The user authority management mechanism is used to set the operation authority of viewing, analyzing, modifying and other operations of the quality inspection results according to different user roles (such as administrators, agents, etc.).

7. The system according to claim 1, characterized in that The appeal processing unit includes an appeal reviewer, which is used for the administrator to review the appeal request and adjust the quality inspection result.

8. The system according to claim 1, characterized in that The data statistics and analysis unit includes a trend prediction analyzer, which is used to predict future quality trends based on historical quality detection analysis data and traffic data, and provide forward-looking support for management decisions.

9. A method for transcribing and analyzing the quality of telephone recordings using the system according to any one of claims 1 to 8, characterized in that: The following steps are involved: 1) Configure the quality inspection template, including setting inspection points and corresponding scores. 2) Connect to the call center system to obtain recordings and related data. 3) Filter recordings and initiate quality inspection tasks. 4) Automatically perform voice recognition and transcribe recordings into text content. 5) Use the transcribed text content for intelligent comparison to generate quality inspection results. 6) View, analyze, and modify quality inspection results in the management interface. 7) Handle the complaint requests of agents and adjust the quality inspection results. 8) Use the statistics and analysis module to visualize the quality inspection analysis data and call data.

10. The method according to claim 11, wherein step 1) further includes customizing the score of the detection point and the associated quality detection label and warning trigger condition through the template editor according to the call quality detection requirements and compliance requirements. Step 4) further includes using advanced speech recognition algorithms and noise reduction algorithms to ensure highly accurate text transcription in complex environments. Step 5) further includes using natural language processing technology to perform semantic analysis on the transcribed text content to enhance the accuracy and depth of the quality detection results. Step 6) further includes using machine learning algorithms to learn the administrator's modification behavior to automatically optimize the quality detection template and detection algorithm. Step 7) further includes using the complaint result data to self-optimize the quality detection template to improve the system's adaptability and detection accuracy. Step 8) further includes using big data analysis technology to deeply mine the quality detection analysis data to reveal hidden data trends and patterns; using cloud computing platforms to perform distributed computing on large-scale data sets to improve the speed and efficiency of data processing.

Citation Information

Patent Citations

  • Offline voice quality inspection and analysis system for call center

    CN111835925A