Electric power customer service system, processing method and computer medium

Abstract

The invention discloses an electric power customer service system, a processing method and a computer medium, and the system comprises an input interface which is used for interacting with a user and receiving a question input by the user; the intelligent agent is used for analyzing the question input by the user, when the question input by the user comprises a picture, the picture is forwarded to the picture processing module to be converted into a text form, and the intelligent agent generates an answer according to the question input by the user or the picture converted into the text form and returns the answer to an input interface for display; and the picture processing module is used for converting a picture part in the question input by the user into a text form. According to the method, through the intelligent agent, the complex electric power problem proposed by the user can be automatically identified and analyzed, and the visual language model is used for processing the query including text and non-text contents, so that comprehensive information retrieval and understanding are realized, and the processing efficiency and processing accuracy of the electric power customer service system are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of artificial intelligence technology and electric power industry service system, and more specifically, to an electric power customer service system, a processing method and a computer medium. Background Art

[0002] In the current power customer service system, traditional automatic answering machines and manual customer service operations are usually used. Traditional customer service systems usually rely on preset scripts and limited response templates to handle customer inquiries. These systems are efficient when handling standard questions, but when encountering complex or non-standard questions, they often face difficulties in identification and processing, and cannot provide accurate or satisfactory answers, resulting in poor customer service experience.

[0003] In addition, most of the existing power customer service systems fail to fully utilize advanced computer vision technology and have a weak ability to analyze and identify the content of images or documents uploaded by customers. For example, when users upload meter readings or faulty equipment via pictures, traditional systems are usually unable to directly identify the valid information in the images and need to rely on manual judgment, which not only reduces processing efficiency but also increases operating costs.

[0004] In international patent literature, for example, US Patent US20120209618A1 proposes a customer service system based on text and basic image recognition. Although the system uses certain image processing technologies, it mainly relies on text information processing, and the recognition and processing of complex images are still limited. In addition, the system does not combine the retrieval-enhanced generation model, and its dialogue generation and context understanding capabilities are limited, which is not enough to handle higher-level queries.

[0005] Therefore, the intelligence and automation of the power customer service system in the existing technology have not yet reached the ideal state, especially in terms of efficiency and accuracy when processing complex images, data tables and unstructured information. It is necessary to develop a more efficient and intelligent power customer service system to solve the limitations of the current system in information processing, multimodal interaction and automated services. Summary of the invention

[0006] One of the purposes of the present invention is to provide an electric power customer service system to solve the problems of low efficiency and inaccurate information processing of the existing electric power customer service system when dealing with a large number of electric power service-related issues; the second purpose of the present invention is to provide a processing method for the electric power customer service system; the third purpose of the present invention is to provide a computer medium.

[0007] In order to solve the above technical problems, the technical solution of the present invention is as follows:

[0008] A first aspect of the present invention provides an electric power customer service system, comprising:

[0009] An input interface, wherein the input interface is used to interact with a user and receive questions input by the user;

[0010] An intelligent agent, the intelligent agent is used to analyze the question input by the user. When the question input by the user includes a picture, the picture is forwarded to the picture processing module for conversion into text form. The intelligent agent generates an answer based on the question input by the user or the picture converted into text form and returns it to the input interface for display;

[0011] The image processing module converts the image part of the question input by the user into text form.

[0012] Among the above-mentioned technical means, intelligent agents can automatically identify and parse complex electricity problems raised by users, and process queries including text and non-text content (such as tables and complex pictures) using visual language models to achieve comprehensive information retrieval and understanding, thereby improving the efficiency and accuracy of information processing of the electricity customer service system when dealing with a large number of electricity service-related issues.

[0013] Furthermore, the user input question received by the input interface includes text information and non-text information, and the non-text information includes pictures and PDF files.

[0014] Furthermore, the intelligent agent is a Transformer-based deep learning model and is pre-trained using training data consisting of power industry conversations.

[0015] Furthermore, it also includes a back-end knowledge base, which is used to store professional knowledge in the field of electricity.

[0016] Furthermore, the intelligent agent is also provided with a retrieval enhancement mechanism, and the intelligent agent uses the retrieval enhancement mechanism to search in the backend knowledge base according to the question input by the user or the picture converted into text form, and generates an answer to return to the input interface for display.

[0017] Furthermore, the image processing module uses convolutional networks and natural language processing technology to identify text and objects in the image, and converts the text and objects in the image into text form.

[0018] A second aspect of the present invention provides a processing method for an electric power customer service system, the processing method being applied to the electric power customer service system, comprising the following steps:

[0019] Receive user input through the input interface;

[0020] The intelligent agent determines whether the question input by the user includes a picture. If the question input by the user includes a picture, the intelligent agent forwards the picture to the picture processing module to convert the picture in the question input by the user into text information.

[0021] The intelligent agent generates an answer based on the question input by the user or the picture in the question input by the user converted into text information and returns it to the input interface for display.

[0022] Furthermore, before the agent determines whether the question input by the user includes a picture, the following steps are also included:

[0023] The intelligent agent determines whether the question input by the user is an electricity question. If so, the intelligent agent determines whether the question input by the user includes a picture; if not, the intelligent agent prompts that it is not an electricity question and returns to the input interface display.

[0024] Furthermore, the agent generates an answer based on the question input by the user or the picture in the question input by the user converted into text information and returns it to the input interface for display, including:

[0025] The intelligent agent searches the backend knowledge base according to the question input by the user or the picture in the question input by the user converted into text information by using the retrieval enhancement mechanism, and generates an answer to return to the input interface for display.

[0026] A second aspect of the present invention provides a computer medium having a computer program stored thereon, and when the computer program is executed by a processor, the processing method of the electric power customer service system is implemented.

[0027] Compared with the prior art, the technical solution of the present invention has the following beneficial effects:

[0028] 1. The power customer service system of the present invention is based on the structure of Agent (intelligent agent)-VLM (visual language model)-RAG (retrieval enhancement generation) to greatly improve the efficiency and accuracy of handling power-related issues. By using Agent intelligent agents, the system can automatically identify and parse complex power issues raised by users, and process queries including text and non-text content (such as tables and complex pictures) through the visual language model of VLM to achieve comprehensive information retrieval and understanding. In addition, the retrieval enhancement generation mechanism of RAG further enhances the quality of answers and the generation of relevant information;

[0029] 2. The implementation of the power customer service system of the present invention significantly reduces the cost of manual customer service. According to preliminary estimates, the use of the power customer service system of the present invention can save about 30% of human resource expenditure. At the same time, the implementation of the system can also improve customer satisfaction and reduce user churn, thereby bringing better revenue performance and reducing customer retention costs;

[0030] 3. The present invention improves the response speed and service quality of public power services through an efficient intelligent customer service system, and promotes the innovation and development of the service model of the power industry. It makes power services faster and more accurate, greatly improves people's life experience and satisfaction, and also helps to promote the application and popularization of intelligent technology in more fields; in actual deployment and testing, the system of the present invention has an average response time of 40% and a query accuracy of 50% when processing power-related queries compared to traditional customer service systems. When processing complex queries containing non-text content, the system's successful recognition rate reaches 85%, which is significantly higher than the processing effect of ordinary OCR technology. In addition, customer satisfaction surveys show that after using the system of the present invention, customer satisfaction has increased by 60%, effectively reducing customers' waiting time and the difficulty of solving problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 A schematic diagram of the structure of an electric power customer service system provided in Embodiment 1 of the present invention;

[0032] Figure 2 A schematic diagram of the structure of an electric power customer service system provided in Embodiment 2 of the present invention;

[0033] Figure 3 A schematic diagram of a process flow of a power customer service system provided in Embodiment 3 of the present invention;

[0034] Figure 4 A schematic diagram of a specific business process of an electric power customer service system provided in Example 3 of the present invention. DETAILED DESCRIPTION

[0035] The drawings are for illustrative purposes only and should not be construed as limiting the present patent;

[0036] In order to better illustrate the present embodiment, some parts in the drawings may be omitted, enlarged or reduced, and do not represent the size of the actual product;

[0037] It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0038] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0039] Example 1

[0040] This embodiment provides an electric power customer service system, such as Figure 1 As shown, including:

[0041] An input interface, wherein the input interface is used to interact with a user and receive questions input by the user;

[0042] An intelligent agent, the intelligent agent is used to analyze the question input by the user. When the question input by the user includes a picture, the picture is forwarded to the picture processing module for conversion into text form. The intelligent agent generates an answer based on the question input by the user or the picture converted into text form and returns it to the input interface for display;

[0043] The image processing module converts the image part of the question input by the user into text form.

[0044] In a further embodiment, the question input by the user received by the input interface includes text information and non-text information, and the non-text information includes pictures and PDF files.

[0045] In a further embodiment, the agent is a Transformer-based deep learning model and is pre-trained using training data consisting of power industry conversations.

[0046] In this embodiment, the agent is designed as the front end of the system and is the component that interacts directly with the user. The agent has an advanced dialogue management system that can parse the user's language input, understand the intention, and generate corresponding answers. The core of the agent is a Transformer-based deep learning model, which has been trained with a large amount of power industry dialogue data to enable it to understand power-related terms and their context.

[0047] In a further embodiment, the image processing module uses convolutional networks and natural language processing technology to identify text and objects in the image, and converts the text and objects in the image into text form.

[0048] Additionally, for content that is difficult to convert directly, such as tables or specific icons, the module uses its visual capabilities to directly mark and identify these specific areas.

[0049] In this embodiment, the image processing module is used to process non-text information uploaded by users, such as images or PDF files. By combining convolutional neural networks (CNN) with natural language processing technology, VLM can recognize text and objects in images and convert them into structured data that the system can understand. For example, it can extract consumption data from power bill images or identify fault types from equipment images.

[0050] Example 2

[0051] This embodiment provides an electric power customer service system, such as Figure 2 As shown, including:

[0052] An input interface, wherein the input interface is used to interact with a user and receive questions input by the user;

[0053] An intelligent agent, the intelligent agent is used to analyze the question input by the user, and when the question input by the user includes a picture, the picture is forwarded to the picture processing module for conversion into text form, and the intelligent agent uses the retrieval enhancement mechanism to search in the back-end knowledge base according to the question input by the user or the picture converted into text form, and generates an answer to return to the input interface for display;

[0054] An image processing module, which converts the image part of the question input by the user into text form;

[0055] A back-end knowledge base, wherein the back-end knowledge base is used to store professional knowledge in the power field.

[0056] In a further embodiment, the question input by the user received by the input interface includes text information and non-text information, and the non-text information includes pictures and PDF files.

[0057] In a further embodiment, the agent is a Transformer-based deep learning model and is pre-trained using training data consisting of power industry conversations.

[0058] In this embodiment, the agent is designed as the front end of the system and is the component that interacts directly with the user. The agent has an advanced dialogue management system that can parse the user's language input, understand the intention, and generate corresponding answers. The core of the agent is a Transformer-based deep learning model, which has been trained with a large amount of power industry dialogue data to enable it to understand power-related terms and their context.

[0059] In a further embodiment, the image processing module uses convolutional networks and natural language processing technology to identify text and objects in the image, and converts the text and objects in the image into text form.

[0060] In this embodiment, the image processing module is used to process non-text information uploaded by users, such as images or PDF files. By combining convolutional neural networks (CNN) with natural language processing technology, VLM can recognize text and objects in images and convert them into structured data that the system can understand. For example, it can extract consumption data from power bill images or identify fault types from equipment images.

[0061] Additionally, for content that is difficult to convert directly, such as tables or specific icons, the module uses its visual capabilities to directly mark and identify these specific areas.

[0062] In a further embodiment, the retrieval enhancement mechanism is responsible for enhancing the accuracy and relevance of answer generation. The system simultaneously utilizes the knowledge-based retrieval mechanism and the pattern-based generation mechanism to retrieve the most relevant information content from the system's knowledge base, and then generates accurate and reasonable answers in combination with the context of the current conversation.

[0063] Example 3

[0064] This embodiment provides a processing method for an electric power customer service system, and the processing method is applied to the electric power customer service system described in Embodiment 1 or Embodiment 2, such as Figure 3 As shown, the following steps are included:

[0065] Receive user input through the input interface;

[0066] The intelligent agent determines whether the question input by the user includes a picture. If the question input by the user includes a picture, the intelligent agent forwards the picture to the picture processing module to convert the picture in the question input by the user into text information.

[0067] The intelligent agent generates an answer based on the question input by the user or the picture in the question input by the user converted into text information and returns it to the input interface for display.

[0068] In a further embodiment, before the agent determines whether the question input by the user includes a picture, the following steps are further included:

[0069] The intelligent agent determines whether the question input by the user is an electricity question. If so, the intelligent agent determines whether the question input by the user includes a picture; if not, the intelligent agent prompts that it is not an electricity question and returns to the input interface display.

[0070] In this embodiment, if Figure 4 As shown, the agent also determines the type of question input by the user:

[0071] If the type of question input by the user is an electricity fee inquiry, the intelligent agent calculates the electricity fee by obtaining the inquiry information and then returns to the input interface for display;

[0072] If the type of question input by the user is a question about the use of an electrical appliance, the agent returns the answer to the input interface after searching the vector database in the backend database;

[0073] If the type of question entered by the user is a complaint, the complaint is sent to the backend for manual processing;

[0074] If the type of question input by the user is a power failure report, after obtaining the user's location, the fault information is sent to the enterprise WeChat backend, waiting for the staff to handle the fault;

[0075] If the type of question input by the user is other questions, the agent calls the search engine to search for the answer and returns the answer input interface for display.

[0076] In a specific embodiment, a user uploaded a picture of his broken electric meter, and the system not only automatically identified the model and degree of damage, but also guided the user on how to submit a repair request. In addition, for another user's high electricity bill inquiry, the system intelligently prompted possible abnormal usage patterns or equipment failures by analyzing the images of the electricity bills uploaded by the user in recent months.

[0077] In a further embodiment, the agent generates an answer based on the question input by the user or the picture in the question input by the user converted into text information and returns it to the input interface for display, including:

[0078] The intelligent agent searches the backend knowledge base according to the question input by the user or the picture in the question input by the user converted into text information by using the retrieval enhancement mechanism, and generates an answer to return to the input interface for display.

[0079] Example 4

[0080] This embodiment provides a computer medium, on which a computer program is stored. When the computer program is executed by a processor, the processing method of the electric power customer service system described in Embodiment 3 is implemented.

[0081] The same or similar reference numerals correspond to the same or similar components;

[0082] The terms used in the drawings to describe positional relationships are only used for illustrative purposes and should not be construed as limiting this patent;

[0083] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the embodiments here. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the claims of the present invention.

Claims

1. An electric power customer service system, characterized in that: include: An input interface, wherein the input interface is used to interact with a user and receive questions input by the user; An intelligent agent, the intelligent agent is used to analyze the question input by the user. When the question input by the user includes a picture, the picture is forwarded to the picture processing module for conversion into text form. The intelligent agent generates an answer based on the question input by the user or the picture converted into text form and returns it to the input interface for display; The image processing module converts the image part of the question input by the user into text form.

2. The power customer service system according to claim 1, characterized in that: The user input question received by the input interface includes text information and non-text information, and the non-text information includes pictures and PDF files.

3. The power customer service system according to claim 1, characterized in that: The agent is a Transformer-based deep learning model and is pre-trained using training data consisting of power industry conversations.

4. The power customer service system according to claim 3, characterized in that: The system also includes a backend knowledge base, which is used to store professional knowledge in the field of electricity.

5. The power customer service system according to claim 4, characterized in that: The intelligent agent is also provided with a search enhancement mechanism. The intelligent agent searches the backend knowledge base according to the question input by the user or the picture converted into text form by using the search enhancement mechanism, and generates an answer to return to the input interface for display.

6. The power customer service system according to claim 1, characterized in that: The image processing module uses convolutional networks and natural language processing technology to identify text and objects in the image, and converts the text and objects in the image into text form.

7. A processing method for an electric power customer service system, characterized in that: The processing method is applied to the power customer service system according to any one of claims 1 to 6, comprising the following steps: Receive user input through the input interface; The intelligent agent determines whether the question input by the user includes a picture. If the question input by the user includes a picture, the intelligent agent forwards the picture to the picture processing module to convert the picture in the question input by the user into text information. The intelligent agent generates an answer based on the question input by the user or the picture in the question input by the user converted into text information and returns it to the input interface for display.

8. The processing method of the electric power customer service system according to claim 7, characterized in that: Before the agent determines whether the question input by the user includes a picture, the following steps are also included: The intelligent agent determines whether the question input by the user is an electricity question. If so, the intelligent agent determines whether the question input by the user includes a picture; if not, the intelligent agent prompts that it is not an electricity question and returns to the input interface display.

9. The processing method of the electric power customer service system according to claim 7, characterized in that: The intelligent agent generates an answer based on the question input by the user or the picture in the question input by the user converted into text information and returns it to the input interface for display, including: The intelligent agent searches the backend knowledge base according to the question input by the user or the picture in the question input by the user converted into text information by using the retrieval enhancement mechanism, and generates an answer to return to the input interface for display.

10. A computer medium, characterized in that The computer medium stores a computer program, and when the computer program is executed by the processor, the processing method of the electric power customer service system according to any one of claims 7 to 9 is implemented.

Images