Summary method and system based on online session, storage medium and equipment

Abstract

The invention provides a summarizing method and system based on an online session, a storage medium and equipment. The summarizing method comprises the following steps: acquiring session content in a current online session group; performing session analysis on the session content, selecting a preset time period, and determining effective session content corresponding to the time period in the session content; calling an AI model to analyze the effective session content to obtain a session summary report; judging the truth rate of the current session summary report; and when the truth rate is smaller than the actual truth rate, correcting the content in the session summary report, so that the corresponding truth rate in the corrected session summary report meets the actual truth rate. According to the method, complex and diversified natural languages can be deeply and accurately understood and analyzed, and correction can be carried out in the final summary report, so that the AI large model can be fed back, and the accuracy of the session summary report is improved.

Description

Technical Field

[0001] The present invention relates to the field of human-computer interaction technology, and in particular to a summary method, system, storage medium and device based on online conversation. Background Art

[0002] Online conversations are a form of communication in the sales industry. In particular, the maintenance and management of customer relationships in the automotive sales industry increasingly relies on online communication tools, such as corporate WeChat and customer WeChat. The amount of information exchanged between customers and dealer consultants is huge and complex and diverse. Effective analysis of these conversations is crucial to improving sales efficiency and customer satisfaction.

[0003] At present, the conversations are usually analyzed and summarized by rule-making systems, simple keyword matching methods, or primary classification models; rule-based systems process conversations by setting a series of fixed rules and patterns in advance. For example, if a customer mentions "discount", it is classified as a price-related issue. However, this method has obvious limitations. The formulation of rules often cannot cover all possible expressions, and the ability to understand complex semantics and contexts is insufficient; methods based on simple keyword matching: only filter and classify conversation content based on pre-set keywords. For example, set keywords such as "configuration" and "color" to identify the focus of customer attention. However, this method cannot accurately understand the customer's true intentions and may lead to misjudgment or omission of important information due to the limitations of keywords; primary classification models based on machine learning: use some basic machine learning algorithms, such as naive Bayes classifiers, to classify conversations. However, these models usually have poor data adaptability and are difficult to accurately identify newly emerging conversation patterns and expressions.

[0004] Therefore, the existing methods of summarizing conversations have problems such as low accuracy, poor adaptability, and inability to deeply understand semantics, making it difficult to efficiently and accurately analyze customer conversation needs. Summary of the invention

[0005] Based on this, the purpose of the present invention is to provide a summary method, system, storage medium and device based on online conversations to solve the technical problem that the existing way of summarizing conversations in the background technology has problems such as low accuracy, poor adaptability, inability to deeply understand semantics, and difficulty in efficiently and accurately analyzing customer conversations.

[0006] A first aspect of the present invention is to provide a summary method based on online conversations, the generation method comprising:

[0007] Get the conversation content in the current online conversation group;

[0008] Performing conversation analysis on the conversation content, selecting a preset time period, and determining valid conversation content in the conversation content corresponding to the time period;

[0009] Calling the AI ​​model to analyze the effective conversation content to obtain a conversation summary report;

[0010] Determine the truthfulness of the current session summary report;

[0011] When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

[0012] Furthermore, before the step of performing conversation analysis on the conversation content, the method further includes:

[0013] Determining the type of the online conversation group;

[0014] When the current online conversation group is a pre-sale conversation group, obtaining a personnel structure in the pre-sale conversation group;

[0015] Determine whether there is a customer assistant in the personnel structure;

[0016] If not, the corresponding customer assistant is added to the pre-sales conversation group.

[0017] Furthermore, after the step of determining the type of the online conversation group, the method further includes:

[0018] When the current online conversation group is an after-sales conversation group, obtaining a personnel structure in the after-sales conversation group;

[0019] Determine whether there is a customer assistant in the personnel structure;

[0020] If not, the corresponding customer assistant is added to the after-sales conversation group.

[0021] Further, after the step of adding the corresponding customer assistant to the after-sales conversation group, the method further includes:

[0022] Obtaining a conversation summary report corresponding to the after-sales conversation group;

[0023] Determining whether there is customer demand information in the conversation summary report;

[0024] If so, a demand summary sub-report and a conversation summary sub-report corresponding to the customer demand are generated.

[0025] Further, after the step of generating a demand summary sub-report and a conversation summary sub-report corresponding to the customer demand, the method further includes:

[0026] On the N+1th day after obtaining the demand sub-summary report, determining whether the demand in the demand prompt information has received demand feedback;

[0027] If not, the demand prompt information is used as a demand prompt so that the demand in the demand prompt information is fed back.

[0028] Furthermore, after the step of providing the demand prompt information with a demand prompt so that the demand in the demand prompt information is fed back, the method further includes:

[0029] The demand feedback information is summarized in the session summary sub-report to determine the corresponding session summary report in the after-sales session group.

[0030] Further, after the step of summarizing the demand feedback information into the session summary sub-report to determine the corresponding session summary report in the after-sales session group, the method further includes:

[0031] Determining the truthfulness of the session summary report;

[0032] When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

[0033] A second aspect of the present invention is to provide a summary system based on online conversations, the system comprising:

[0034] The acquisition module is used to obtain the conversation content in the current online conversation group;

[0035] An analysis module performs conversation analysis on the conversation content, selects a preset time period, and determines valid conversation content in the conversation content corresponding to the time period;

[0036] Calling an AI module, which is used to call an AI model to analyze the valid conversation content to obtain a conversation summary report;

[0037] A judgment module, for judging the authenticity of the current session summary report;

[0038] When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

[0039] A third aspect of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, which implements the above-mentioned online conversation-based summarization method when executed by a processor.

[0040] A fourth aspect of the present invention is to provide a summary device based on an online conversation, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned summary method based on the online conversation when executing the program.

[0041] Compared with the prior art, the online conversation-based summary method, system, storage medium and device shown in the present invention have the following beneficial effects:

[0042] In a summary method based on online conversations provided by the present invention, by first obtaining the conversation content in the current online conversation group, and performing conversation analysis on the conversation content, and selecting a preset time period, determining the valid conversation content corresponding to the time period in the conversation content, and then calling the AI ​​model to analyze the valid conversation content to obtain a conversation summary report, judging the authenticity of the current conversation summary report, when the authenticity is less than the actual authenticity, correcting the content in the conversation summary report so that the corresponding authenticity in the corrected conversation summary report meets the actual authenticity, thereby introducing the AI ​​model in the application of the actual online chat scene with customers, it is possible to deeply and accurately understand and analyze complex and diverse natural languages, adapt to the expression habits and contexts of various customers, assist in improving the understanding of the real intentions of customers by front-line personnel, accurately judge the resolution of customer demands, and generate targeted and practical to-do items, and can be corrected in the final summary report, so as to feed back the AI ​​big model, improve the accuracy of the conversation summary report, and solve the problems of low accuracy, poor adaptability, inability to deeply understand semantics, and difficulty in efficiently and accurately analyzing customer conversations in the existing technology for conversation summarization. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 is a flowchart of a summary method based on online conversations in a first embodiment of the present invention;

[0044] Figure 2 A schematic diagram of the structure of an acquisition module in a third embodiment of the present invention;

[0045] Figure 3 Schematic diagram of a summary device based on online conversations in a fourth embodiment of the present invention.

[0046] The following specific implementation manner will further illustrate the present invention in conjunction with the above-mentioned drawings. DETAILED DESCRIPTION

[0047] In order to facilitate the understanding of the present invention, the present invention will be described more fully below with reference to the relevant drawings. Several embodiments of the present invention are given in the drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive.

[0048] It should be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "left", "right" and similar expressions used herein are for illustrative purposes only.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items.

[0050] First embodiment

[0051] See also Figure 1 , which shows a summary method based on online conversation in the first embodiment of the present invention, and the generation method includes:

[0052] S01, obtaining the conversation content in the current online conversation group.

[0053] In this embodiment, the online conversation group can be a WeChat conversation group for communicating with customers, a conversation group between WeChat and enterprise WeChat, a QQ conversation group, etc. It should be noted that in this embodiment, the online conversation group is not limited to the above-mentioned WeChat conversation group, the conversation group between WeChat and enterprise WeChat, and the QQ conversation group, it can also include other types of online conversation groups. Since other online conversation groups belong to the conventional existing technology in this field, they are not specifically explained here.

[0054] S02, performing conversation analysis on the conversation content, selecting a preset time period, and determining valid conversation content in the conversation content corresponding to the time period.

[0055] Specifically, in this embodiment, the preset time period reads the conversation content with the current customer within two hours by default. However, in some actual situations, when the conversation content in the first hour is valid conversation content, and there is no valid conversation content in the second hour, but there is valid conversation content in the third hour, the preset time period can be manually modified to read all valid conversation content. Correspondingly, it also supports summarizing some valid conversation content within the time range.

[0056] S03, calling the AI ​​model to analyze the effective conversation content to obtain a conversation summary report;

[0057] It should be noted that the AI ​​model is an existing technology. The AI ​​model has strong learning and generalization capabilities, can adapt to ever-changing conversation patterns and new ways of expression, ensure accurate analysis and processing of customer conversations, and provide dealer consultants with more valuable information and guidance. It can solve the shortcomings of some primary machine learning models, and utilize the AI ​​model's powerful language understanding and reasoning capabilities to adaptively handle various complex customer expressions and context changes, reduce information omissions and misjudgments, and more accurately summarize customer problems and demands. At the same time, the AI ​​model can also get rid of excessive reliance on preset keywords through in-depth semantic understanding and context analysis, accurately capture the customer's true intentions, and improve the accuracy of understanding customer problems and the reliability of judgment of demands.

[0058] S04, judging the authenticity rate of the current session summary report, when the authenticity rate is less than the actual authenticity rate, proceeding to step S05, when the authenticity rate is not less than the actual authenticity rate, the current session summary report is the final correct session summary report.

[0059] Specifically, in step S05, when the truth rate is less than the actual truth rate, the content of the conversation summary report is corrected so that the corresponding truth rate in the corrected conversation summary report meets the actual truth rate.

[0060] It should be noted that, in this embodiment, the content in the conversation summary report may be manually edited and corrected.

[0061] Specifically, in this example, in order to improve the authenticity and accuracy of the conversation summary report output by the AI ​​model, during the early use process, the conversation summary report output by the AI ​​model is manually corrected so that the corrected conversation summary report can be used for deep learning by the AI ​​model, thereby making the next conversation summary report more authentic.

[0062] In summary, the online conversation summary method shown in this embodiment has at least the following beneficial effects compared with the online conversation summary method in the prior art:

[0063] In a summary method based on online conversations provided by the present invention, by first obtaining the conversation content in the current online conversation group, and performing conversation analysis on the conversation content, and selecting a preset time period, determining the valid conversation content corresponding to the time period in the conversation content, and then calling the AI ​​model to analyze the valid conversation content to obtain a conversation summary report, judging the authenticity of the current conversation summary report, when the authenticity is less than the actual authenticity, correcting the content in the conversation summary report so that the corresponding authenticity in the corrected conversation summary report meets the actual authenticity, thereby introducing the AI ​​model in the application of the actual online chat scene with customers, it is possible to deeply and accurately understand and analyze complex and diverse natural languages, adapt to the expression habits and contexts of various customers, assist in improving the understanding of the real intentions of customers by front-line personnel, accurately judge the resolution of customer demands, and generate targeted and practical to-do items, and can be corrected in the final summary report, so as to feed back the AI ​​big model, improve the accuracy of the conversation summary report, and solve the problems of low accuracy, poor adaptability, inability to deeply understand semantics, and difficulty in efficiently and accurately analyzing customer conversations in the existing technology for conversation summarization.

[0064] Second embodiment

[0065] The second embodiment of the present invention provides a summary method based on an online conversation. The summary method based on an online conversation in this embodiment is different from the summary method based on an online conversation in the first embodiment in that, in this embodiment:

[0066] Before the step of performing conversation analysis on the conversation content, it also includes:

[0067] Determine the type of online conversation group;

[0068] When the current online conversation group is a pre-sales conversation group, obtain the personnel structure in the pre-sales conversation group;

[0069] Determine whether there are customer assistants in the personnel structure;

[0070] If not, the corresponding customer assistant will be added to the pre-sales conversation group.

[0071] It should be noted that the customer assistant is an important position responsible for establishing and maintaining good relationships with customers. It requires good communication skills, service awareness, sales ability and teamwork spirit. It has related docking work such as new customer development, customer relationship maintenance, sales support, order processing, etc.

[0072] Furthermore, after the step of determining the type of the online conversation group, the method further includes:

[0073] When the current online conversation group is an after-sales conversation group, obtain the personnel structure in the after-sales conversation group;

[0074] Determine whether there are customer assistants in the personnel structure;

[0075] If not, the corresponding customer assistant will be added to the after-sales conversation group.

[0076] It should be noted that, in this embodiment, the pre-sales conversation group and the after-sales conversation group are different forms of work content docking, especially in the after-sales conversation group, which has customer complaints.

[0077] Specifically, after the step of adding the corresponding customer assistant to the after-sales conversation group, the method further includes:

[0078] Obtain a conversation summary report corresponding to the after-sales conversation group;

[0079] Determine whether there is customer demand information in the conversation summary report;

[0080] If so, a demand summary sub-report and a conversation summary sub-report corresponding to the customer demand are generated.

[0081] On the N+1th day after obtaining the demand sub-summary report, determine whether the demand in the demand prompt information has received demand feedback;

[0082] If not, the demand prompt information is used as a demand prompt so that the demand in the demand prompt information is fed back.

[0083] Summarize the demand feedback information into the conversation summary sub-report to determine the corresponding conversation summary report in the after-sales conversation group and determine the authenticity rate of the conversation summary report;

[0084] When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

[0085] It should be noted that the above-mentioned day N+1 is the second day after the sub-summary report of the appeal is obtained.

[0086] In summary, the online conversation-based summarization method shown in this embodiment has at least the following beneficial effects compared with the online conversation-based summarization method in the first embodiment:

[0087] In the summary method based on online conversations in the second embodiment of the present invention, by first judging the type of the current online conversation group, it is possible to accurately output a conversation summary report on whether it contains customer demands. When there are customer demands in the conversation summary report, a demand prompt can be made on the N+1th day of obtaining the demand sub-summary report, so that real-time feedback can be provided on the customer's demands, and the final feedback results are summarized in the conversation summary sub-report, so that a conversation summary report with solutions to customer demands is finally obtained.

[0088] Third embodiment

[0089] Another aspect of the present invention is to provide a summary system based on online conversations. Figure 2 , which shows a summary system based on online conversations in a third embodiment of the present invention, and the summary system based on online conversations includes:

[0090] An acquisition module 11 is used to acquire the conversation content in the current online conversation group;

[0091] The analysis module 12 is used to perform conversation analysis on the conversation content, select a preset time period, and determine the valid conversation content corresponding to the time period in the conversation content;

[0092] Calling the AI ​​module 13, which is used to call the AI ​​model to analyze the effective conversation content to obtain a conversation summary report;

[0093] A judgment module 14, used to judge the authenticity of the current session summary report;

[0094] When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

[0095] Fourth embodiment

[0096] Another aspect of the present invention also provides a summary device based on online conversations, see Figure 3 , shown is a summary device based on online conversations in the fourth embodiment of the present invention, including a memory 20, a processor 10, and a computer program 30 stored in the memory and executable on the processor. When the processor 10 executes the computer program 30, the summary method based on online conversations as described above is implemented.

[0097] Among them, the device based on the online conversation summary method can be specifically hardware, and the processor 10 in some embodiments can be a central processing unit (CPU), a controller, a microcontroller, a microprocessor or other data processing chip, which is used to run the program code stored in the memory 20 or process data, such as executing access restriction programs.

[0098] The memory 20 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 20 can be an internal storage unit of the summary device based on the online session, such as the hard disk of the summary device based on the online session. In other embodiments, the memory 20 can also be an external storage device of the summary device based on the online session, such as a plug-in hard disk equipped on the summary device based on the online session, a smart memory card (SmartMediaCard, SMC), a secure digital (SecureDiɡital, SD) card, a flash card (FlashCard), etc. Further, the memory 20 can also include both an internal storage unit and an external storage device of the summary device based on the online session. The memory 20 can be used not only to store application software and various types of data installed in the summary device based on the online session, but also to temporarily store data that has been output or is to be output.

[0099] It should be pointed out that Figure 3 The illustrated structure does not constitute a limitation on the online conversation-based summary device. In other embodiments, the online conversation-based summary device may include fewer or more components than those illustrated, or may combine certain components, or arrange the components differently.

[0100] The embodiment of the present invention further provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the online conversation-based summarization method as described above is implemented.

[0101] Those skilled in the art will appreciate that the logic or steps represented in the flowchart or otherwise described herein, for example, may be considered as an ordered list of executable instructions for implementing logical functions, and may be specifically implemented in any computer-readable storage medium for use by an instruction execution system, device or apparatus (e.g., based on a computer system, including a processor system or other system that can fetch instructions from an instruction execution system, device or apparatus and execute an instruction system), or in conjunction with such instruction execution systems, devices or apparatuses. For purposes of this specification, a "computer-readable storage medium" may be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0102] More specific examples of computer-readable storage media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable storage medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or processing in another suitable manner as necessary, and then stored in a computer memory.

[0103] It should be understood that the various parts of the present invention can be implemented in hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or a combination thereof: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PgA), a field programmable gate array (FpgA), etc.

[0104] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0105] The above embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present invention. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the attached claims.

Claims

1. A summary method based on online conversation, characterized in that: The summarizing method comprises: Get the conversation content in the current online conversation group; Performing conversation analysis on the conversation content, selecting a preset time period, and determining valid conversation content in the conversation content corresponding to the time period; Calling the AI ​​model to analyze the effective conversation content to obtain a conversation summary report; Determine the truthfulness of the current session summary report; When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

2. The online conversation-based summarization method according to claim 1, characterized in that: Before the step of performing conversation analysis on the conversation content, the method further includes: Determining the type of the online conversation group; When the current online conversation group is a pre-sale conversation group, obtaining a personnel structure in the pre-sale conversation group; Determine whether there is a customer assistant in the personnel structure; If not, the corresponding customer assistant is added to the pre-sales conversation group.

3. The online conversation-based summarization method according to claim 2, characterized in that: After the step of determining the type of the online conversation group, the method further includes: When the current online conversation group is an after-sales conversation group, obtaining a personnel structure in the after-sales conversation group; Determine whether there is a customer assistant in the personnel structure; If not, the corresponding customer assistant is added to the after-sales conversation group.

4. The online conversation-based summarization method according to claim 3, characterized in that: After the step of adding the corresponding customer assistant to the after-sales conversation group, the method further includes: Obtaining a conversation summary report corresponding to the after-sales conversation group; Determining whether there is customer demand information in the session summary report; If so, a demand summary sub-report and a conversation summary sub-report corresponding to the customer demand are generated.

5. The online conversation-based summarization method according to claim 4, characterized in that: After the step of generating a demand summary sub-report and a conversation summary sub-report corresponding to the customer demand, the method further includes: On the N+1th day after obtaining the demand sub-summary report, determining whether the demand in the demand prompt information has received demand feedback; If not, the demand prompt information is used as a demand prompt so that the demand in the demand prompt information is fed back.

6. The online conversation-based summarization method according to claim 5, characterized in that: After the step of providing the demand prompt information with a demand prompt so that the demand in the demand prompt information is fed back, the method further includes: The demand feedback information is summarized in the session summary sub-report to determine the corresponding session summary report in the after-sales session group.

7. The online conversation-based summarization method according to claim 6, characterized in that: After the step of summarizing the demand feedback information into the session summary sub-report to determine the corresponding session summary report in the after-sales session group, the method further includes: Determining the truthfulness of the session summary report; When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

8. A summary system based on online conversations, characterized in that: The system is used to implement the online conversation-based summarization method according to any one of claims 1 to 7, and the system comprises: The acquisition module is used to obtain the conversation content in the current online conversation group; An analysis module performs conversation analysis on the conversation content, selects a preset time period, and determines valid conversation content in the conversation content corresponding to the time period; Calling an AI module, which is used to call an AI model to analyze the valid conversation content to obtain a conversation summary report; A judgment module, for judging the authenticity of the current session summary report; When the truth rate is less than the actual truth rate, the content of the session summary report is corrected so that the corresponding truth rate in the corrected session summary report meets the actual truth rate.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the online conversation-based summarization method according to any one of claims 1 to 7 is implemented.

10. A summary device based on online conversations, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the online conversation-based summarization method according to any one of claims 1 to 7 is implemented.

Images