Work order automatic extracting and summarizing method and system based on real-time conversation

By collecting and automatically processing call content in real time, and generating work orders using voice recognition and machine learning algorithms, the problems of low work order entry efficiency and inaccurate information extraction are solved, efficient and accurate work order processing is achieved, and customer service quality is improved .

CN118917790APending Publication Date: 2024-11-08BEIYIN FINANCIAL TECH CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410959559.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

In the prior art, work order entry efficiency is low and information extraction is inaccurate, and it is susceptible to human factors, resulting in information omission or misunderstanding, affecting service quality and efficiency.

Method used

The automatic extraction and summary method based on real-time calls is adopted, and the call content is collected in real time, and voice recognition technology is used to convert it into text, pre-process and key information extraction are performed, and work orders are generated and summarized in combination with machine learning algorithms.

Benefits of technology

Real-time and accurate work order entry and information extraction are achieved, manual operations are reduced, work efficiency and information integrity are improved, error risks caused by human factors are reduced, and the quality and efficiency of customer service are improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118917790A_ABST
    Figure CN118917790A_ABST
Patent Text Reader

Abstract

The invention discloses a work order automatic extraction and summarization method and system based on real-time conversation. The method comprises the following steps: acquiring conversation content with a client in real time; converting the collected voice data into text data; carrying out preprocessing on the converted text data; key information is extracted from the preprocessed text; generating a work order according to the extracted key information, and filling the work order according to the key information; and summarizing the generated work order by using a machine learning algorithm. According to the method and the system, efficient and accurate service requirements can be met, the situation that a large number of seats overflow does not occur especially in the service peak period, and the service quality and the user experience are ensured.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent analysis, and specifically relates to a method and system for automatically extracting and summarizing work orders based on real-time calls. Background Art

[0002] With the rapid development of banking business and the geometric growth of customer service needs, a large amount of consulting work has exploded, with characteristics such as uncertain service time, large number of visits, highly professional questions, and prominent personalized issues.

[0003] The main existing technical solution is to manually query, record, and extract key information in the work order system during or after the call between the agent and the customer, mainly including the caller number, work order number, work order type, problem category, problem description, and handling situation description. The work order system needs to enter information in different fields, and then submit the work order approval process, which will be circulated to the business nodes of each processor. After the work order is completed, the work order status will be changed to form a closed loop of the business process.

[0004] In the process of data entry, manual data entry is significantly less efficient. This traditional method relies on staff to manually enter data one by one, which is not only time-consuming and labor-intensive, but also easily affected by human factors such as fatigue and distraction. At the same time, manual data entry also requires a large number of paper or electronic documents as support, which undoubtedly increases resource consumption and storage costs, resulting in unnecessary waste of resources.

[0005] In terms of information processing, the accuracy of information extraction is also a problem faced by manual methods. For example, during a call, staff need to listen carefully and manually extract key information. However, due to the limited hearing and memory ability of humans, this manual extraction method often cannot guarantee the integrity and accuracy of information. In addition, the human ear also has certain limitations in distinguishing different accents, speaking speeds and intonations, which further increases the risk of information errors.

[0006] Manual methods also have limitations in summarizing information. Workers need to screen, classify and summarize a large amount of information. This process is not only time-consuming and laborious, but also prone to information loss or omission due to human factors. In addition, due to differences in everyone's understanding ability and way of thinking, different workers may come to different conclusions or summaries even for the same piece of information. This uncertainty makes it difficult to ensure the comprehensiveness and accuracy of the results of manual summarization.

[0007] More importantly, in the process of manually extracting work orders, human factors have become a risk point that cannot be ignored. Due to negligence, misunderstanding or fatigue of staff, key information in the work order may be omitted, misunderstood or recorded incorrectly. These errors will not only affect the subsequent workflow, but may also lead to serious consequences such as customer complaints and service delays. Therefore, in the work order processing, the manual extraction method often has certain risks and instability.

[0008] It is difficult to meet the requirements of efficient and accurate service by using manual seats alone to provide work order recording services. Especially during peak service periods, it leads to a large number of seat overflows, which seriously affects the service quality and user experience. Summary of the invention

[0009] In view of the above problems, the present invention is proposed to provide a method and system for automatically extracting and summarizing work orders based on real-time calls, which overcome the above problems or at least partially solve the above problems.

[0010] To achieve the above object, the present invention adopts the following technical solutions:

[0011] A method for automatically extracting and summarizing work orders based on real-time calls, the method comprising:

[0012] Collect conversation content with customers in real time;

[0013] Convert the collected voice data into text data;

[0014] Preprocess the converted text data;

[0015] Extract key information from the preprocessed text;

[0016] Generate a work order based on the extracted key information and fill in the work order based on the key information;

[0017] Use machine learning algorithms to summarize the generated work orders.

[0018] Optionally, use a recording device or API interface to collect the conversation content with the customer in real time.

[0019] Optionally, speech recognition technology is used to convert the collected speech data into text data.

[0020] Optionally, the speech recognition technology introduces a recurrent neural network model and a long short-term memory network model.

[0021] Optionally, the preprocessing includes cleaning, word segmentation and part-of-speech tagging.

[0022] Optionally, the extraction of key information from the preprocessed text uses named entity recognition technology and relationship extraction technology.

[0023] Optionally, the key information includes the work order number, problem description and urgency.

[0024] Optionally, the machine learning algorithm includes a classification algorithm and a clustering algorithm.

[0025] Optionally, the machine learning algorithm trains a work order processing model by learning and analyzing historical work order data.

[0026] The present invention also provides a system for automatically extracting and summarizing work orders based on real-time calls, using the method for automatically extracting and summarizing work orders based on real-time calls as described above, comprising:

[0027] The call content collection module collects the call content with customers in real time;

[0028] The speech recognition module converts the collected speech data into text data;

[0029] A text preprocessing module preprocesses the converted text data;

[0030] Key information extraction module, extracts key information from the preprocessed text;

[0031] The work order automatic extraction module generates a work order based on the extracted key information and fills the work order based on the key information;

[0032] The work order summary module uses machine learning algorithms to summarize the generated work orders.

[0033] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are:

[0034] 1. The present invention has a real-time transcription function, which can instantly convert the conversation between the user and the agent into text form when the user is talking to the agent, and accurately enter it into the work order system. This function greatly improves work efficiency and reduces the time and error rate of manual entry. At the same time, real-time transcription can also ensure the immediacy and accuracy of information, providing reliable data support for subsequent work order processing.

[0035] 2. The present invention also has a powerful information extraction capability. Once the call content is transcribed into text, the tool will automatically analyze the text information and extract key information according to preset rules and algorithms. This key information includes but is not limited to user name, contact information, problem description, etc., and will be automatically filled in the corresponding fields in the work order system. This function not only improves the accuracy of information entry, but also greatly reduces the workload of manual operation, making the entire work order processing process more efficient and automated.

[0036] 3. The present invention also has the function of automatic summarization. Based on the transcription and extraction of information, the tool will conduct in-depth analysis and processing of this information, and automatically extract the key points and key information of the call. Then, the results of these summaries will be automatically filled into the corresponding fields in the work order system, providing a more comprehensive and accurate reference for subsequent work order processing. This function not only makes work order processing more convenient and efficient, but also effectively avoids information omissions or misunderstandings caused by human factors, and improves the quality and efficiency of the entire customer service. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A flowchart of a method for automatically extracting and summarizing work orders based on real-time calls provided in an embodiment of the present application. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0039] Embodiment 1:

[0040] See also Figure 1 The first embodiment provides a method for automatically extracting and summarizing work orders based on real-time calls, including:

[0041] S1. Collect conversation content with customers in real time.

[0042] Use recording equipment or API interfaces to collect conversation content with customers in real time.

[0043] S2. Convert the collected voice data into text data.

[0044] The collected voice data is converted into text data using speech recognition technology. In order to improve the accuracy and speed of speech recognition, advanced deep learning algorithms are used. Specifically, speech recognition technology introduces recurrent neural network models and long short-term memory network models.

[0045] These models have strong memory and learning capabilities, can process complex voice signals, and accurately identify the content. By continuously optimizing and adjusting model parameters, accurate voice recognition is achieved, providing users with a more convenient and efficient interactive experience.

[0046] S3. Preprocess the converted text data.

[0047] The preprocessing includes operations such as cleaning, word segmentation and part-of-speech tagging to improve the accuracy of subsequent analysis.

[0048] S4, extract key information from the preprocessed text;

[0049] Named entity recognition and relationship extraction technologies are used to extract key information from preprocessed text. These technologies can quickly and accurately extract key information from text, such as names of people, places, time, etc. In view of the special needs of the banking field, a large number of professional vocabulary and contextual information are also combined to optimize the natural language processing model. Through these technologies, users' intentions and needs can be better understood, providing strong support for subsequent work order processing.

[0050] The key information includes work order number, problem description, urgency and other information.

[0051] S5. Generate a work order based on the extracted key information, and fill in the work order based on the key information.

[0052] S6. Use machine learning algorithms to summarize the generated work orders.

[0053] Using machine learning algorithms, the generated work orders are summarized and summarized to identify common problem types, customer demand trends, etc., providing decision support for banks.

[0054] The machine learning algorithm includes a classification algorithm and a clustering algorithm. The machine learning algorithm trains a set of efficient work order processing models by learning and analyzing a large amount of historical work order data. The model can automatically identify common problem types, customer demand trends, etc., and provide targeted suggestions for work order processing. This not only improves the efficiency of work order processing, but also reduces the risk of human error, providing users with more professional and accurate services.

[0055] In order to achieve seamless connection between various modules, voice recognition, natural language processing, machine learning and other modules are integrated to form a complete set of intelligent devices. At the same time, in order to facilitate the connection with the bank's existing customer service system, work order management system, etc., a rich API interface is also provided. These interfaces support a variety of data formats and communication protocols, and can meet the connection requirements of different systems. Through system integration and interface connection, the seamless connection between intelligent devices and the bank's existing systems is achieved, providing the bank with more comprehensive and efficient customer service support.

[0056] This embodiment has a real-time transcription function, which can instantly convert the conversation between the user and the agent into text form when the user is talking to the agent, and accurately enter it into the work order system. This function greatly improves work efficiency and reduces the time and error rate of manual entry. At the same time, real-time transcription can also ensure the immediacy and accuracy of information, providing reliable data support for subsequent work order processing.

[0057] This embodiment also has a powerful information extraction capability. Once the call content is transcribed into text, the tool will automatically analyze the text information and extract key information according to preset rules and algorithms. This key information includes but is not limited to the user's name, contact information, problem description, etc., and will be automatically filled in the corresponding fields in the work order system. This function not only improves the accuracy of information entry, but also greatly reduces the workload of manual operations, making the entire work order processing process more efficient and automated.

[0058] This embodiment also has the function of automatic summarization. Based on the transcription and extraction of information, the tool will conduct in-depth analysis and processing of this information, and automatically extract the key points and key information of the call. Then, the results of these summaries will be automatically filled in the corresponding fields in the work order system, providing a more comprehensive and accurate reference for subsequent work order processing. This function not only makes work order processing more convenient and efficient, but also effectively avoids information omissions or misunderstandings caused by human factors, and improves the quality and efficiency of the entire customer service.

[0059] Embodiment 2:

[0060] Embodiment 2 provides a work order automatic extraction and summarization system based on real-time calls, using a work order automatic extraction and summarization method based on real-time calls as provided in Embodiment 1, including:

[0061] Call content collection module, collects the call content with customers in real time;

[0062] The speech recognition module converts the collected speech data into text data;

[0063] A text preprocessing module preprocesses the converted text data;

[0064] Key information extraction module, extracts key information from the preprocessed text;

[0065] The work order automatic extraction module generates a work order based on the extracted key information and fills the work order based on the key information;

[0066] The work order summary module uses machine learning algorithms to summarize the generated work orders.

[0067] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A method for automatically extracting and summarizing work orders based on real-time calls, characterized in that: The method comprises: Collect conversation content with customers in real time; Convert the collected voice data into text data; Preprocess the converted text data; Extract key information from the preprocessed text; Generate a work order based on the extracted key information and fill in the work order based on the key information; Use machine learning algorithms to summarize the generated work orders.

2. A method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 1, characterized in that: Use recording equipment or API interfaces to collect conversation content with customers in real time.

3. A method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 1, characterized in that: The collected voice data is converted into text data using speech recognition technology.

4. A method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 3, characterized in that: The speech recognition technology introduces a recurrent neural network model and a long short-term memory network model.

5. The method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 1, characterized in that: The preprocessing includes cleaning, word segmentation and part-of-speech tagging.

6. The method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 1, characterized in that: The extraction of key information from the preprocessed text uses named entity recognition technology and relationship extraction technology.

7. The method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 1, characterized in that: The key information includes the work order number, problem description and urgency.

8. The method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 1, characterized in that: The machine learning algorithms include classification algorithms and clustering algorithms.

9. The method for automatically extracting and summarizing work orders based on real-time calls as claimed in claim 1, characterized in that: The machine learning algorithm trains a work order processing model by learning and analyzing historical work order data.

10. A work order automatic extraction and summary system based on real-time calls, characterized in that: A method for automatically extracting and summarizing work orders based on real-time calls according to any one of claims 1 to 9, comprising: The call content collection module collects the call content with customers in real time; The speech recognition module converts the collected speech data into text data; A text preprocessing module preprocesses the converted text data; Key information extraction module, extracts key information from the preprocessed text; The work order automatic extraction module generates a work order based on the extracted key information and fills the work order based on the key information; The work order summary module uses machine learning algorithms to summarize the generated work orders.

Citation Information

Cited By

  • Work order generation method and device based on natural language processing, equipment and medium

    CN119761334A