Customer service knowledge base optimization method and device based on large model and medium

Abstract

The invention discloses a customer service knowledge base optimization method and device based on a large model and a medium. The method comprises the following steps: acquiring a to-be-added business document to be added to a customer service knowledge base, and segmenting the to-be-added business document into a plurality of business texts suitable for large model processing length; performing key information extraction on the plurality of service texts by utilizing the large model, and obtaining a plurality of service features corresponding to the plurality of service texts; utilizing a large model to generate a plurality of potential questions and answers corresponding to the plurality of business texts according to the plurality of business features corresponding to the plurality of business texts; combining the plurality of potential questions and answers corresponding to the plurality of business texts by using a large model to form a plurality of pieces of customer service knowledge, and adding the formed plurality of pieces of customer service knowledge to a customer service knowledge base to obtain an updated customer service knowledge base; and carrying out customer service knowledge sampling inspection on the updated customer service knowledge base, and issuing the updated customer service knowledge base under the condition that a sampling inspection result is qualified.

Description

Technical Field

[0001] The present invention relates to the technical field of customer service knowledge base optimization, and more specifically, to a customer service knowledge base optimization method, device and medium based on a large model. Background Art

[0002] With the rapid development of information technology and the Internet, intelligent customer service systems have been widely used in all walks of life. Traditional customer service knowledge bases are mainly built and maintained manually, and there are several problems in use:

[0003] 1. Low construction efficiency

[0004] Traditional customer service knowledge base construction mainly relies on manual work, which extracts useful information from massive documents, records and materials based on experience, and then manually organizes and compiles it into knowledge base content. This method is not only time-consuming and labor-intensive, taking up a lot of man-hours of front-line professional employees, but is also prone to omissions and errors, resulting in the knowledge base being incomplete and inaccurate.

[0005] 2. Difficulty in optimization and maintenance

[0006] In actual applications, the customer service knowledge base needs to be continuously updated and optimized to cope with new problems and new needs. However, manually updating and optimizing the knowledge base also requires a lot of manpower. Especially for large enterprises and organizations, the knowledge base is large in scale and updated frequently, making optimization and maintenance more difficult. This may lead to untimely maintenance, affecting the response speed of the customer service system and user experience. In addition, the inevitable subjectivity and inconsistency of manual operations may also lead to unstable knowledge base quality. Summary of the invention

[0007] In view of the deficiencies in the prior art, the present invention provides a customer service knowledge base optimization method, device and medium based on a large model.

[0008] According to one aspect of the present invention, a method for optimizing a customer service knowledge base based on a large model is provided, comprising:

[0009] Obtain the business document to be added to the customer service knowledge base, and divide the business document to be added into multiple business texts of a length suitable for processing by the large model;

[0010] Use the big model to extract key information from multiple business texts and obtain multiple business features corresponding to the multiple business texts;

[0011] Use the big model to generate multiple potential questions and multiple answers corresponding to multiple business texts based on multiple business features corresponding to multiple business texts;

[0012] Use the big model to combine multiple potential questions and answers corresponding to multiple business texts to form multiple customer service knowledge, and add the multiple customer service knowledge to the customer service knowledge base to obtain an updated customer service knowledge base;

[0013] Conduct a random check of customer service knowledge on the updated customer service knowledge base, and if the random check results are qualified, publish the updated customer service knowledge base.

[0014] Optionally, it also includes:

[0015] Collect data on unsatisfactory questions and unanswered questions in the daily Q&A process of the customer service knowledge base at preset intervals;

[0016] A large model is used to optimize the corresponding customer service knowledge in the customer service knowledge base based on the data of dissatisfied questions and unanswered questions.

[0017] Optionally, service feature F i The expression is:

[0018] F i =FeatureExtract(T i )

[0019] Where, T i The i-th business text segmented from the business document D to be added, D = T1 + T2 + ... T n , i=1,2,...,n; F i is the business feature of the i-th business text;

[0020] Potential Problems i The expression is:

[0021] Q i = QuestionGenerate(F i )

[0022] In the formula, Q i is the potential problem of the i-th business text;

[0023] Answer A i The expression is:

[0024] A i =AnswerGenerarte(F i )

[0025] In the formula, A i is the answer to the i-th business text.

[0026] Optionally, a large model is used to combine multiple potential questions and answers corresponding to multiple business texts to form multiple customer service knowledge, including:

[0027] Integrate multiple potential issues of multiple business documents to determine the potential issue set Q = Q1 + Q2 + ... Q n ;

[0028] Integrate multiple answers to multiple business documents to determine the answer set A=A1+A2+...A n ;

[0029] Combine potential question sets and answer sets to form multiple customer service knowledge QAs.

[0030] According to another aspect of the present invention, there is provided a customer service knowledge base optimization device based on a large model, comprising:

[0031] A segmentation module is used to obtain the business document to be added to the customer service knowledge base, and segment the business document to be added into multiple business texts of a length suitable for processing by a large model;

[0032] An extraction module is used to extract key information from multiple business texts using a large model to obtain multiple business features corresponding to the multiple business texts;

[0033] A generation module, used to generate multiple potential questions and multiple answers corresponding to multiple business texts based on multiple business features corresponding to multiple business texts using a large model;

[0034] A formation module is used to combine multiple potential questions and answers corresponding to multiple business texts using a large model to form multiple customer service knowledge, and add the formed multiple customer service knowledge to a customer service knowledge base to obtain an updated customer service knowledge base;

[0035] The sampling module is used to perform a sampling check on the customer service knowledge of the updated customer service knowledge base, and if the sampling check results are qualified, the updated customer service knowledge base is released.

[0036] Optionally, it also includes:

[0037] The collection module is used to collect data on unsatisfactory questions and unanswered questions in the daily Q&A process of the customer service knowledge base at preset intervals;

[0038] The optimization module is used to optimize the corresponding customer service knowledge in the customer service knowledge base according to the dissatisfied question data and the unanswered question data using a large model.

[0039] Optionally, service feature F i The expression is:

[0040] F i =FeatureExtract(T i )

[0041] Where, T i The i-th business text segmented from the business document D to be added, D = T1 + T2 + ... T n , i=1,2,...,n; F i is the business feature of the i-th business text;

[0042] Potential Problems i The expression is:

[0043] Q i = QuestionGenerate(F i )

[0044] In the formula, Q i is the potential problem of the i-th business text;

[0045] Answer A i The expression is:

[0046] A i =AnswerGenerate(F i )

[0047] In the formula, A i is the answer to the i-th business text.

[0048] Optionally, the formation module uses a large model to combine multiple potential questions and answers corresponding to multiple business texts to form multiple customer service knowledge, including:

[0049] The first integration submodule is used to integrate multiple potential problems of multiple business documents to determine a potential problem set Q = Q1 + Q2 + ... Q n ;

[0050] The second integration submodule is used to integrate multiple answers to multiple business documents to determine the answer set A=A1+A2+...A of the business document to be added n ;

[0051] The combination submodule is used to combine potential question sets and answer sets to form multiple customer service knowledge QAs.

[0052] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the storage medium stores a computer program, and the computer program is used to execute the method described in any one of the above aspects of the present invention.

[0053] According to another aspect of the present invention, an electronic device is provided, comprising: a processor; a memory for storing instructions executable by the processor; the processor is configured to read the executable instructions from the memory and execute the instructions to implement the method described in any one of the above aspects of the present invention.

[0054] The present invention has the following advantages:

[0055] 1. Improve efficiency: Use large models to automatically generate question-answer pairs, expand similar questions and rewrite answers, which significantly improves the speed of knowledge base construction.

[0056] 2. Ensure quality: The large model can generate high-quality consistent content based on business documents and user historical data, reducing quality problems caused by manual differences.

[0057] 3. Easy to expand: Through automated processing, the expansion and maintenance of the knowledge base becomes more convenient, reducing the tediousness and errors of manual operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] A more complete understanding of exemplary embodiments of the present invention may be obtained by referring to the following drawings:

[0059] Figure 1 It is a flow chart of a customer service knowledge base optimization method based on a large model provided by an exemplary embodiment of the present invention;

[0060] Figure 2 It is another flow chart of a customer service knowledge base optimization method based on a large model provided by an exemplary embodiment of the present invention;

[0061] Figure 3 It is a structural schematic diagram of a customer service knowledge base optimization device based on a large model provided by an exemplary embodiment of the present invention;

[0062] Figure 4 This is a structure of an electronic device provided by an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0063] Below, the exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments of the present invention, and it should be understood that the present invention is not limited to the exemplary embodiments described here.

[0064] It should be noted that the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless specifically stated otherwise.

[0065] Those skilled in the art can understand that the terms "first" and "second" in the embodiments of the present invention are only used to distinguish different steps, devices or modules, etc., and neither represent any specific technical meaning nor indicate the necessary logical order between them.

[0066] It should also be understood that, in the embodiments of the present invention, “plurality” may refer to two or more than two, and “at least one” may refer to one, two or more than two.

[0067] It should also be understood that any component, data or structure mentioned in the embodiments of the present invention can generally be understood as one or more, unless explicitly limited or otherwise indicated in the context.

[0068] In addition, the term "and / or" in the present invention is only a description of the association relationship of the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in the present invention generally indicates that the associated objects before and after are in an "or" relationship.

[0069] It should also be understood that the description of the various embodiments of the present invention focuses on the differences between the various embodiments, and the same or similar aspects thereof can be referenced to each other, and for the sake of brevity, they will not be described one by one.

[0070] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0071] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0072] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0073] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0074] Embodiments of the present invention can be applied to electronic devices such as terminal devices, computer systems, servers, etc., which can operate with many other general or special computing system environments or configurations. Examples of well-known terminal devices, computing systems, environments and / or configurations suitable for use with electronic devices such as terminal devices, computer systems, servers, etc. include, but are not limited to: personal computer systems, server computer systems, thin clients, thick clients, handheld or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, small computer systems, large computer systems, and distributed cloud computing technology environments including any of the above systems, etc.

[0075] Electronic devices such as terminal devices, computer systems, servers, etc. can be described in the general context of computer system executable instructions (such as program modules) executed by computer systems. Generally, program modules can include routines, programs, object programs, components, logic, data structures, etc., which perform specific tasks or implement specific abstract data types. Computer systems / servers can be implemented in a distributed cloud computing environment, where tasks are performed by remote processing devices linked through a communication network. In a distributed cloud computing environment, program modules can be located on local or remote computing system storage media including storage devices.

[0076] Exemplary Methods

[0077] Figure 1 FIG. 1 is a flow chart of a method for optimizing a customer service knowledge base based on a large model provided by an exemplary embodiment of the present invention. This embodiment can be applied to electronic devices, such as Figure 1 As shown, the customer service knowledge base optimization method 100 based on the large model includes the following steps:

[0078] Step 101, obtaining a business document to be added to the customer service knowledge base, and dividing the business document to be added into multiple business texts of a length suitable for processing by a large model;

[0079] Step 102, using the large model to extract key information from multiple business texts respectively, and obtaining multiple business features corresponding to the multiple business texts;

[0080] Step 103, using the large model to generate multiple potential questions and multiple answers corresponding to the multiple business texts according to the multiple business features corresponding to the multiple business texts;

[0081] Step 104: using the big model to combine multiple potential questions and answers corresponding to multiple business texts to form multiple customer service knowledge, and adding the formed multiple customer service knowledge to the customer service knowledge base to obtain an updated customer service knowledge base;

[0082] Step 105: Perform a customer service knowledge spot check on the updated customer service knowledge base, and if the spot check result is qualified, publish the updated customer service knowledge base.

[0083] Specifically, in order to improve the efficiency of customer service knowledge base construction and optimization and reduce maintenance costs, large model technology can be used to automatically construct and optimize the customer service knowledge base.

[0084] The solution of the present invention changes the traditional way of constructing the knowledge base of intelligent customer service from manual compilation by business experts to large-scale model processing using artificial intelligence technology.

[0085] Users (U) conduct Q&A and consultation through the intelligent customer service Q&A interface (CHAT). CHAT completely relies on the pre-built domain knowledge base (KB). KB consists of multiple question-answer pairs (QA). It is recalled through various methods such as keywords and semantic retrieval to select the knowledge and answers that best meet the user's intentions. These are returned to U through the CHAT component, realizing a complete customer service consultation Q&A process. Therefore, the accuracy, completeness, and timeliness of the knowledge base directly affect the Q&A effect.

[0086] In the original knowledge base construction process, business documents (D) will be directly constructed by knowledge administrators (KBA) and then manually updated to KB. Because the cost of business learning and organization is high, time-consuming and labor-intensive, it will lead to untimely updates and poor question-answering results. The present invention optimizes the construction and update of the knowledge base with the help of large model technology. Figure 2 As shown, the specific implementation process is as follows:

[0087] The business document (D) is divided into several texts T1, T2...Tn according to the processing length of the large model, D = T1+T2+...Tn. Use the large model service (LLM) to extract large sections of text from T, represented by Q,A = ExtractQA(T). This process is mainly divided into three stages: feature extraction, question generation, and answer generation:

[0088] Feature extraction: extract key information from each sub-segment Ti to form a feature representation Fi, using F i =FeatureExtract(T i )express.

[0089] Question generation, based on the feature representation Fi, generates the corresponding potential question Qi. i = QuestionGenerate(F i )express.

[0090] Answer generation, based on the feature representation Fi, generates the corresponding answer Ai. Using formula A i =AnswerGenerate(Fi )express.

[0091] Finally, all generated questions and answers are combined to obtain a set of question and answer pairs:

[0092] Q=Q1+Q2+...Qn, A=A1+A2+...An, finally forming a QA pair and building it into a knowledge base.

[0093] In order to ensure the quality of the constructed knowledge base (KB), KBA needs to conduct random inspections or quality inspections. Only after confirmation can it be finally put into the warehouse and released to serve the online customer service system.

[0094] At the same time, in daily question-and-answer data, two types of questions worth optimizing are collected: dissatisfied questions (DQ) and unanswered questions (UQ).

[0095] The unsatisfactory question is that U thinks that the intelligent customer service’s response to his question is irrelevant, so he gave a bad review. This shows that our knowledge base may have defects for this problem, or the question is missing or ambiguous;

[0096] Unanswered questions are questions asked by users that cannot be answered in the knowledge base, so the intelligent customer service has to give a fallback reply, indicating that there must be a missing question and the knowledge base needs to be supplemented and improved.

[0097] Both types of problems are obvious points for optimization of the knowledge base, and optimization is needed to improve the quality of the KB.

[0098] In the traditional process: DQ and UQ issues will be fed back to KBA. KBA will adjust and optimize the corresponding QA in KB based on its business domain knowledge. It takes about 10 minutes to answer a DQ or UQ question. If the modification is not timely when the daily Q&A volume is large, DQ and UQ issues will accumulate more and more, and the entire intelligent customer service system will fall into a vicious circle.

[0099] In the optimized process: DQ and UQ issues will be directly fed back to LLM. LLM will adjust and optimize QA based on its business domain knowledge. This process is carried out in batches, and one day's data can be processed within 10 minutes. KBA only needs to review and verify, which takes less than one minute per item on average. While effectively reducing time costs, it improves the quality of KB and even the quality of customer service questions and answers.

[0100] In the workflow of knowledge base construction, optimization of unsatisfactory questions, and optimization of unanswered questions, the most time-consuming QA extraction and QA construction work are separated and handed over to the big model for processing. Business experts only review the results, which ultimately improves the overall work efficiency and ensures the quality of the knowledge base.

[0101] This method has the following significant advantages:

[0102] 1. Dynamic Optimization

[0103] The large model can automatically update and optimize the knowledge base content based on the latest business data set documents. Through continuous learning and adjustment, the model can continuously improve the coverage of the knowledge base and the accuracy of the answers, ensuring that the knowledge base is always in the best state.

[0104] 2. Reduce costs

[0105] The automation capabilities of large models significantly reduce reliance on manual operations, reduce human resource investment and maintenance costs. Enterprises can invest more resources in other core businesses and improve overall operational efficiency and competitiveness.

[0106] 3. Improve user experience

[0107] Through intelligent optimization of large models, the customer service knowledge base can provide more accurate, fast and consistent answers, improving user experience and satisfaction. The intelligent customer service system can better deal with complex and changing problems and enhance user stickiness.

[0108] The key technical points of the present invention are:

[0109] Apply big model technology to the construction and optimization of FAQ knowledge base, and construct QA question-answer pairs through text segmentation, feature extraction, question generation, and answer generation;

[0110] From business knowledge documents to unanswered questions and unsatisfactory questions online, knowledge construction and optimization are processed in batches using big model technology, which improves processing efficiency;

[0111] In order to ensure the accuracy and authority of the knowledge base, the knowledge base information optimized by large model technology is manually reviewed, which ensures the quality of questions and answers while improving overall efficiency.

[0112] The present invention has the following advantages:

[0113] 1. Improve efficiency: Use large models to automatically generate question-answer pairs, expand similar questions and rewrite answers, which significantly improves the speed of knowledge base construction.

[0114] 2. Ensure quality: The large model can generate high-quality consistent content based on business documents and user historical data, reducing quality problems caused by manual differences.

[0115] 3. Easy to expand: Through automated processing, the expansion and maintenance of the knowledge base becomes more convenient, reducing the tediousness and errors of manual operations.

[0116] Exemplary Devices

[0117] Figure 3 FIG. 1 is a schematic diagram of a structure of a customer service knowledge base optimization device based on a large model provided by an exemplary embodiment of the present invention. Figure 3 As shown, the device 300 includes:

[0118] A segmentation module 310 is used to obtain a business document to be added to the customer service knowledge base, and segment the business document to be added into multiple business texts of a length suitable for processing by a large model;

[0119] An extraction module 320 is used to extract key information from multiple business texts using a large model to obtain multiple business features corresponding to the multiple business texts;

[0120] A generation module 330, for generating a plurality of potential questions and a plurality of answers corresponding to the plurality of business texts by using the large model according to a plurality of business features corresponding to the plurality of business texts;

[0121] A forming module 340 is used to combine multiple potential questions and answers corresponding to multiple business texts using a large model to form multiple customer service knowledge, and add the formed multiple customer service knowledge to a customer service knowledge base to obtain an updated customer service knowledge base;

[0122] The sampling module 350 is used to perform a sampling test on the updated customer service knowledge base, and publish the updated customer service knowledge base if the sampling test result is qualified.

[0123] Optionally, the apparatus 300 further includes:

[0124] The collection module is used to collect data on unsatisfactory questions and unanswered questions in the daily Q&A process of the customer service knowledge base at preset intervals;

[0125] The optimization module is used to optimize the corresponding customer service knowledge in the customer service knowledge base according to the dissatisfied question data and the unanswered question data using a large model.

[0126] Optionally, service feature F i The expression is:

[0127] F i =FeatureExtract(T i )

[0128] Where, T i The i-th business text segmented from the business document D to be added, D = T1 + T2 + ... T n , i=1,2,...,n; F i is the business feature of the i-th business text;

[0129] Potential Problems iThe expression is:

[0130] Q i = QuestionGenerate(F i )

[0131] In the formula, Q i is the potential problem of the i-th business text;

[0132] Answer A i The expression is:

[0133] A i =AnswerGenerate(F i )

[0134] In the formula, A i is the answer to the i-th business text.

[0135] Optionally, the forming module 340 uses a large model to combine multiple potential questions and answers corresponding to multiple business texts to form multiple customer service knowledge, including:

[0136] The first integration submodule is used to integrate multiple potential problems of multiple business documents to determine a potential problem set Q = Q1 + Q2 + ... Q n ;

[0137] The second integration submodule is used to integrate multiple answers to multiple business documents to determine the answer set A=A1+A2+...A to be added to the business document n ;

[0138] The combination submodule is used to combine potential question sets and answer sets to form multiple customer service knowledge QAs.

[0139] Exemplary Electronic Devices

[0140] Figure 4 This is a structure of an electronic device provided by an exemplary embodiment of the present invention. Figure 4 As shown, the electronic device 40 includes one or more processors 41 and a memory 42 .

[0141] The processor 41 may be a central processing unit (CPU) or other forms of processing units having data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device to perform desired functions.

[0142] The memory 42 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory (cache), etc. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 41 may run the program instructions to implement the methods of the software programs of the various embodiments of the present invention described above and / or other desired functions. In one example, the electronic device may also include: an input device 43 and an output device 44, which are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0143] In addition, the input device 43 may also include, for example, a keyboard, a mouse, etc.

[0144] The output device 44 can output various information to the outside, and can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto.

[0145] Of course, to simplify, Figure 4 Only some of the components related to the present invention in the electronic device are shown, and components such as a bus, an input / output interface, etc. are omitted. In addition, the electronic device may further include any other appropriate components according to specific application conditions.

[0146] Exemplary computer program products and computer-readable storage media

[0147] In addition to the above-mentioned methods and devices, an embodiment of the present invention may also be a computer program product, which includes computer program instructions, which, when executed by a processor, enable the processor to execute the steps of the method according to various embodiments of the present invention described in the above-mentioned "Exemplary Method" section of this specification.

[0148] The computer program product may be written in any combination of one or more programming languages ​​to write program code for performing the operations of the embodiments of the present invention, including object-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0149] In addition, an embodiment of the present invention may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, enable the processor to execute the steps of the method according to various embodiments of the present invention described in the above “Exemplary Method” section of this specification.

[0150] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can include, for example, but is not limited to, a system, system or device of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination of the above. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0151] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details disclosed above are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0152] Each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the system embodiment, since it basically corresponds to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0153] The block diagrams of the devices, systems, equipment, and systems involved in the present invention are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, systems, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The word "such as" used here refers to the phrase "such as but not limited to", and can be used interchangeably with it.

[0154] The method and system of the present invention may be implemented in many ways. For example, the method and system of the present invention may be implemented by software, hardware, firmware or any combination of software, hardware, firmware. The above order of steps for the method is only for illustration, and the steps of the method of the present invention are not limited to the order specifically described above, unless otherwise specifically stated. In addition, in some embodiments, the present invention may also be implemented as a program recorded in a recording medium, which includes machine-readable instructions for implementing the method according to the present invention. Thus, the present invention also covers a recording medium storing a program for executing the method according to the present invention.

[0155] It should also be noted that in the system, device and method of the present invention, each component or each step can be decomposed and / or recombined. These decompositions and / or recombinations should be regarded as equivalent schemes of the present invention. The above description of the disclosed aspects is provided to enable any technician in the field to make or use the present invention. Various modifications to these aspects are very obvious to those skilled in the art, and the general principles defined here can be applied to other aspects without departing from the scope of the present invention. Therefore, the present invention is not intended to be limited to the aspects shown here, but in accordance with the widest range consistent with the principles and novel features disclosed here.

[0156] The above description has been given for the purpose of illustration and description. In addition, this description is not intended to limit the embodiments of the present invention to the forms disclosed herein. Although a number of example aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, changes, additions and sub-combinations thereof.

Claims

1. A customer service knowledge base optimization method based on a large model, characterized in that: include: Acquire a business document to be added to the customer service knowledge base, and divide the business document to be added into multiple business texts of a length suitable for processing by a large model; Utilizing the large model to extract key information from multiple business texts respectively, and obtaining multiple business features corresponding to the multiple business texts; Using the large model to generate multiple potential questions and multiple answers corresponding to the multiple business texts according to the multiple business features corresponding to the multiple business texts; Using the large model, multiple potential questions and answers corresponding to multiple business texts are combined to form multiple customer service knowledge, and the formed multiple customer service knowledge is added to the customer service knowledge base to obtain an updated customer service knowledge base; Conduct a customer service knowledge spot check on the updated customer service knowledge base, and if the spot check results are qualified, publish the updated customer service knowledge base.

2. The method according to claim 1, characterized in that: Also includes: Collecting data on unsatisfactory questions and unanswered questions in the daily Q&A process of the customer service knowledge base at preset intervals; The large model is used to optimize the corresponding customer service knowledge in the customer service knowledge base according to the dissatisfied question data and the unanswered question data.

3. The method according to claim 1, characterized in that The service feature F i The expression is: F i =FeatureExtract(T i ) Where, T i The i-th business text segmented from the business document D to be added, D = T1 + T2 + ... T n , i=1,2,...,n; F i is the business feature of the i-th business text; The potential problem Q i The expression is: Q i =QuestionGenerate(F i ) In the formula, Q i is the potential problem of the i-th business text; The answer A i The expression is: A i =AnswerGenerate(F i ) In the formula, A i is the answer to the i-th business text.

4. The method according to claim 3, characterized in that: The large model is used to combine multiple potential questions and answers corresponding to multiple business texts to form multiple customer service knowledge, including: Integrate multiple potential problems of multiple business documents to determine the potential problem set Q=Q1+Q2+...Q n ; Integrate multiple answers to multiple business documents to determine the answer set A=A1+A2+...A of the business document to be added n ; The potential question set and the answer set are combined to form multiple customer service knowledge QAs.

5. A customer service knowledge base optimization device based on a large model, characterized in that: include: A segmentation module, used for acquiring a business document to be added to the customer service knowledge base, and segmenting the business document to be added into a plurality of business texts of a length suitable for processing by a large model; An extraction module, used to extract key information from multiple business texts respectively using the large model to obtain multiple business features corresponding to the multiple business texts; A generation module, used to generate a plurality of potential questions and a plurality of answers corresponding to a plurality of business texts by using the large model according to a plurality of business features corresponding to a plurality of business texts; A forming module, used to combine multiple potential questions and answers corresponding to multiple business texts using the large model to form multiple customer service knowledge, and add the formed multiple customer service knowledge to the customer service knowledge base to obtain an updated customer service knowledge base; The sampling module is used to perform a sampling test on the updated customer service knowledge base, and if the sampling test result is qualified, publish the updated customer service knowledge base.

6. The device according to claim 5, characterized in that Also includes: A collection module, used to collect data on unsatisfactory questions and unanswered questions in the daily question-and-answer process of the customer service knowledge base at preset intervals; An optimization module is used to optimize the corresponding customer service knowledge in the customer service knowledge base according to the dissatisfied question data and the unanswered question data using the large model.

7. The device according to claim 5, characterized in that The service feature F i The expression is: F i =FeatureExtract(T i ) Where, T i The i-th business text segmented from the business document D to be added, D = T1 + T2 + ... T n , i=1,2,...,n; F i is the business feature of the i-th business text; The potential problem Q i The expression is: Q i =QuestionGenerate(F i ) In the formula, Q i is the potential problem of the i-th business text; The answer A i The expression is: A i =AnswerGenerarte(F i ) In the formula, A i is the answer to the i-th business text.

8. The device according to claim 7, characterized in that In the formation module, the large model is used to combine multiple potential questions and answers corresponding to multiple business texts to form multiple customer service knowledge, including: The first integration submodule is used to integrate multiple potential problems of multiple business documents to determine the potential problem set Q=Q1+Q2+...Q of the business document to be added n ; The second integration submodule is used to integrate multiple answers to multiple business documents to determine the answer set A=A1+A2+...A of the business document to be added n ; The combination submodule is used to combine the potential question set and the answer set to form multiple customer service knowledge QAs.

9. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, and the computer program is used to execute the method according to any one of claims 1 to 4.

10. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the method described in any one of claims 1 to 4.

Images