Dialogue data processing method, dialogue data processing device and related product
By using game theory analysis of dialogue data, the complex tasks of large language models are broken down into multiple intermediate tasks, which improves the stability and accuracy of dialogue data processing and solves the problem of insufficient stability and accuracy of large language models in complex tasks.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-24
- Publication Date
- 2026-03-31
AI Technical Summary
Large language models struggle to maintain stability and accuracy when dealing with complex problems, especially in long-chain reasoning tasks, where existing technologies cannot effectively improve the quality of their arithmetic, common sense, and reasoning.
By introducing game theory analysis, dialogue data is split into multi-round dialogue data for game theory analysis. By obtaining game theory analysis results and strategy information, the stability of task processing and feature extraction effect are improved.
By using intermediate structure processing, the stability and accuracy of dialogue data processing are improved, which can truly reflect the game process and thinking of the two parties in the dialogue and improve the accuracy of reasoning.
Smart Images
Figure CN121766449A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to the field of data processing technology, and in particular to a dialogue data processing method, dialogue data processing apparatus, electronic device, computer-readable storage medium, and computer program product. Background Technology
[0002] Large Language Models (LLMs) are deep learning models trained on large amounts of text data, enabling them to generate natural language text or understand the meaning of language text. When using these LLMs to handle complex problems, they often struggle to directly provide the correct answer. To address this, Chain of Thought (CoT) can be introduced. This method requires or prompts the model to explicitly output intermediate reasoning steps before providing the final answer, thereby enhancing the arithmetic, common sense, and reasoning quality of the large model and improving its performance on complex reasoning tasks. However, when using the Chain of Thought approach, its stability decreases as the chain length increases, and the accuracy of reasoning cannot be guaranteed. Summary of the Invention
[0003] This disclosure provides a dialogue data processing method, a dialogue data processing apparatus, an electronic device, a computer-readable storage medium, and a computer program product.
[0004] In a first aspect, this disclosure provides a dialogue data processing method, which includes: acquiring multi-round dialogue data between a first user and a second user for processing preset business, wherein the multi-round dialogue data is composed of alternating first dialogue data and second dialogue data, the first dialogue data corresponding to the first user and the second dialogue data corresponding to the second user; performing game analysis on each round of dialogue data to obtain a game analysis result corresponding to each round of dialogue data; determining a target question category corresponding to each round of first dialogue data based on the game analysis result corresponding to each round of first dialogue data; for any round of first dialogue data, determining the strategy information adopted by the second user for the first dialogue data based on the second dialogue data associated with the first dialogue data and the game analysis result corresponding to the associated second dialogue data; and obtaining a question-strategy correspondence result based on the question categories of the multi-round first dialogue data and the strategy information corresponding to each round of first dialogue data.
[0005] Secondly, this disclosure provides a dialogue data processing apparatus, comprising: a first acquisition module, configured to acquire multi-round dialogue data between a first user and a second user for processing preset services, wherein the multi-round dialogue data is composed of alternating first dialogue data and second dialogue data, the first dialogue data corresponding to the first user and the second dialogue data corresponding to the second user; an analysis module, configured to perform game analysis on each round of dialogue data to obtain a game analysis result corresponding to each round of dialogue data; a first determination module, configured to determine a target question category corresponding to each round of first dialogue data based on the game analysis result corresponding to each round of first dialogue data; a second determination module, configured to determine, for any round of first dialogue data, the strategy information adopted by the second user for the first dialogue data based on the second dialogue data associated with the first dialogue data and the game analysis result corresponding to the associated second dialogue data; and a second acquisition module, configured to obtain a question-strategy correspondence result based on the question categories of the multi-round first dialogue data and the strategy information corresponding to each round of first dialogue data.
[0006] Thirdly, this disclosure provides an electronic device comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores one or more computer programs executable by the at least one processor, the one or more computer programs being executed by the at least one processor to enable the at least one processor to perform the above-described dialogue data processing method.
[0007] Fourthly, this disclosure provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor / processor core, implements the above-described dialogue data processing method.
[0008] Fifthly, this disclosure provides a computer program product including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code, wherein when the computer-readable code is run in a processor of an electronic device, the processor in the electronic device performs the above-described dialogue data processing method.
[0009] The embodiments provided in this disclosure introduce game theory processing into the processing of preset business operations. Game theory analysis is performed on each round of dialogue data to obtain a game theory analysis result corresponding to each round of dialogue data. This game theory analysis result is an intermediate structure in the reasoning process, which is equivalent to breaking down a large-scale task into multiple intermediate tasks. The corresponding intermediate structure is obtained through the processing of intermediate tasks. This processing method can improve the stability of task processing on the one hand, and fully extract features from the initial multi-round dialogue data on the other hand, thereby obtaining an "information-rich" intermediate structure that can truly reflect the game process and game thinking of the two parties in the dialogue. Furthermore, the target question category corresponding to each round of first dialogue data can be determined based on the game theory analysis result corresponding to each round of first dialogue data. Then, for any round of first dialogue data, the strategy information adopted by the second user for the first dialogue data can be determined based on the second dialogue data associated with the first dialogue data and the game theory analysis result corresponding to the associated second dialogue data. Finally, the question strategy corresponding result can be obtained based on the question category of the multi-round first dialogue data and the strategy information corresponding to each round of first dialogue data. Since an intermediate structure is introduced in the processing, not only can stability be improved, but the feature extraction effect can also be enhanced, thereby improving the accuracy of reasoning.
[0010] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this disclosure, nor is it intended to limit the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description
[0011] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the embodiments of the present disclosure to explain the disclosure and do not constitute a limitation thereof. The above and other features and advantages will become more apparent to those skilled in the art from the detailed description of exemplary embodiments with reference to the accompanying drawings, in which:
[0012] Figure 1 This diagram illustrates an application scenario of the dialogue data processing method and apparatus provided in the embodiments of this disclosure.
[0013] Figure 2 This is a flowchart of a dialogue data processing method provided in an embodiment of the present disclosure.
[0014] Figure 3 This is a schematic diagram illustrating the processing procedure of a dialogue data processing method provided in an embodiment of this disclosure.
[0015] Figure 4 This is a block diagram of a dialogue data processing apparatus provided in an embodiment of the present disclosure.
[0016] Figure 5This is a block diagram of an electronic device provided in an embodiment of the present disclosure. Detailed Implementation
[0017] To enable those skilled in the art to better understand the technical solutions of this disclosure, exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments of this disclosure to aid understanding. These should be considered merely exemplary. Therefore, those skilled in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.
[0018] Where there is no conflict, the various embodiments of this disclosure and the features thereof in the embodiments may be combined with each other.
[0019] As used herein, the term “and / or” includes any and all combinations of one or more related enumerated entries.
[0020] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit this disclosure. As used herein, the singular forms “a” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will also be understood that when the terms “comprising” and / or “made of” are used in this specification, the presence of the stated feature, integral, step, operation, element, and / or component is specified, but the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or groups thereof is not excluded. Words such as “connected” or “linked” are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect.
[0021] Unless otherwise specified, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art. It will also be understood that terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant art and this disclosure, and will not be interpreted as having an idealized or overly formal meaning, unless expressly so defined herein.
[0022] The collection, storage, use, processing, transmission, provision, and disclosure of user personal information in this technical solution comply with relevant laws and regulations and do not violate public order and good morals. The use of user data in this technical solution follows relevant national laws and regulations (e.g., the "Information Security Technology - Personal Information Security Specification"). For example, appropriate measures are taken for personal information access control; restrictions are imposed on the display of personal information; the purpose of using personal information does not exceed the scope of direct or reasonable association; and explicit identity targeting is eliminated when using personal information to avoid precisely identifying specific individuals.
[0023] The reasoning process of large models is inherently unstable. For example, in the non-thinking version of a large model, task reasoning involves internalizing the thinking process; in other words, the information needed for thinking is all contained in the hidden vectors of the model structure. This approach is suitable for simple tasks but struggles with complex ones. In the thinking version of a large model, the thinking process can be output, significantly influencing the final result. Furthermore, since generating data is essentially a process of modeling textual data—equivalent to deriving the conditional probability of the following text based on the preceding context—the large model may not be able to extract all the necessary information during the thinking process. Even a single token difference can lead to a significant change in the final result. All of these factors contribute to the poor performance of large models in tasks requiring long-term thinking.
[0024] In view of the above, embodiments of this disclosure provide a dialogue data processing method, a dialogue data processing apparatus, an electronic device, a computer-readable storage medium, and a computer program product.
[0025] The dialogue data processing method according to the embodiments of this disclosure can improve the stability of processing dialogue data, enhance the feature extraction effect, and thus improve the accuracy of reasoning.
[0026] The dialogue data processing method according to embodiments of this disclosure can be executed by an electronic device such as a terminal device or a server. The terminal device can be an in-vehicle device, user equipment (UE), mobile device, user terminal, terminal, cellular phone, cordless phone, personal digital assistant (PDA), handheld device, computing device, in-vehicle device, wearable device, etc. The method can be implemented by a processor calling computer-readable program instructions stored in memory. Alternatively, the method can be executed by a server.
[0027] In a first aspect, embodiments of this disclosure provide a method for processing dialogue data.
[0028] Figure 1 This diagram illustrates an application scenario of the dialogue data processing method and apparatus provided in the embodiments of this disclosure.
[0029] like Figure 1 As shown, the application scenario of this disclosure embodiment may include terminal device 101, network 103, and server 102. Network 103 is used as a medium to provide a communication link between terminal device 101 and server 102. Network 103 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.
[0030] Users can use terminal device 101 to interact with server 102 via network 103 to receive or send messages, etc. Various communication client applications can be installed on terminal device 101, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (for example only).
[0031] Terminal device 101 can be various electronic devices with a display screen and support web browsing, including but not limited to smartphones, tablets, laptops, and desktop computers.
[0032] Server 102 can be a server that provides various services, such as a backend management server that supports websites browsed by users using terminal device 101 (for example only). The backend management server can analyze and process data such as received user requests, and feed back the processing results (such as web pages, information, or data obtained or generated according to user requests) to the terminal device.
[0033] It should be noted that the dialogue data processing method and apparatus provided in this disclosure embodiment can be executed by server 102. Accordingly, the dialogue data processing method and apparatus provided in this disclosure embodiment can be located in server 102. The dialogue data processing method and apparatus provided in this disclosure embodiment can also be executed by a server or server cluster that is different from server 102 and is capable of communicating with terminal device 101 and / or server 102. Accordingly, the dialogue data processing method and apparatus provided in this disclosure embodiment can also be located in a server or server cluster that is different from server 102 and is capable of communicating with terminal device 101 and / or server 102.
[0034] It should be understood that Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, any number of terminal devices, networks, and servers can be included.
[0035] Figure 2 A flowchart illustrating a dialogue data processing method provided in an embodiment of this disclosure. (Refer to...) Figure 2 The dialogue data processing method may include the following steps.
[0036] Step S21: Obtain multi-round dialogue data between the first user and the second user for processing preset business. The multi-round dialogue data consists of alternating first dialogue data and second dialogue data, with the first dialogue data corresponding to the first user and the second dialogue data corresponding to the second user.
[0037] Step S22: Perform game analysis on the dialogue data for each round to obtain the game analysis results corresponding to each round of dialogue data;
[0038] Step S23: Based on the game analysis results corresponding to the first dialogue data in each round, determine the target question category corresponding to the first dialogue data in each round;
[0039] Step S24: For any round of first dialogue data, based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data, determine the strategy information adopted by the second user for the first dialogue data.
[0040] Step S25: Based on the question categories of the first dialogue data in multiple rounds and the strategy information corresponding to each round of the first dialogue data, obtain the corresponding result of the question strategy.
[0041] In some optional embodiments, the preset service is the service to be processed, and the first user and the second user are the relevant personnel involved in the preset service.
[0042] For example, a pre-set business could be a sales business, where the first user is a potential customer and the second user is a salesperson.
[0043] For example, the preset service could be a consultation service, where the first user is the customer who initiates the consultation request, and the second user is the customer service personnel who provide the response.
[0044] It should be noted that the above examples of preset services, first users, and second users are merely illustrative and are not intended to limit the scope of this disclosure.
[0045] In some optional embodiments, the original dialogue data between the first user and the second user processing preset services needs to be further segmented into dialogue rounds to obtain multi-round dialogue data. This multi-round dialogue data consists of alternating first and second dialogue data, with the first dialogue data corresponding to the first user and the second dialogue data corresponding to the second user. This method divides the dialogue data into small-granularity processing units, facilitating subsequent processing.
[0046] For example, the original dialogue data between the first user and the second user can be split into dialogue slices [1], [1-2], [1-3], ..., [1-n], thus obtaining dialogue data for n+1 rounds, where n>1. Each dialogue slice corresponds to one round of dialogue data, and each round of dialogue data corresponds to either the first user or the second user.
[0047] Furthermore, considering that the first user and the second user may have multiple rounds of dialogue during the processing of preset services, in order to improve the stability and accuracy of the processing, a game analysis processing method is introduced in this embodiment. On the one hand, a long reasoning process can be broken down into multiple shorter game processes, thereby improving the stability of reasoning. On the other hand, multiple game processes can also enhance the feature extraction capability of dialogue data, thereby improving the accuracy of the processing results.
[0048] In some optional embodiments, game analysis is performed on each round of dialogue data to obtain game analysis results corresponding to each round of dialogue data, including: performing game analysis on each round of dialogue data separately to obtain game analysis results corresponding to each round of dialogue data; or, for the i-th round of dialogue data, performing game analysis based on the i-1 game analysis results corresponding to the i-th round of dialogue data and the previous i-1 rounds of dialogue data to obtain game analysis results corresponding to the i-th round of dialogue data, where i>1.
[0049] For example, for the i-th round of dialogue data, game analysis is performed based on the ij game analysis results corresponding to the i-th round of dialogue data and the previous ij rounds of dialogue data, to obtain the game analysis result corresponding to the i-th round of dialogue data, where i>1, 1<j<i.
[0050] Therefore, in each game analysis, we can perform game analysis only on the dialogue data of the current round to obtain the game analysis result corresponding to the dialogue data of the current round; or, we can combine the game analysis results of the dialogue data of several rounds before the current round with the dialogue data of the current round to perform game analysis together to obtain the game analysis result corresponding to the dialogue data of the current round.
[0051] For example, there are N rounds of dialogue data between the first user and the second user. Game analysis is performed on the i-th round of dialogue data to obtain the game analysis result of the i-th round of dialogue data, where 1≤i≤N.
[0052] This approach involves a smaller amount of data processing and correspondingly lower processing complexity, thus allowing for convenient acquisition of the relevant game theory analysis results.
[0053] For example, there are N rounds of dialogue data between the first user and the second user. When performing game analysis on the i-th round of dialogue data, the game analysis is performed based on the i-th round of dialogue data and the i-1 game analysis results corresponding to the 1st to i-1th rounds of dialogue data, so as to obtain the game analysis result of the i-th round of dialogue data, 1≤i≤N.
[0054] In this approach, in addition to the dialogue data of the current round, the game analysis also includes the game analysis results of historical dialogue data. Therefore, historical game information can be fully utilized to obtain more accurate and reasonable game analysis results.
[0055] In some alternative embodiments, the model can be used to perform the corresponding data processing for game analysis. Since the model can participate in multiple inferences after being trained once, game analysis can be performed conveniently and efficiently.
[0056] In some optional embodiments, game analysis is performed on each round of dialogue data to obtain the game analysis result corresponding to each round of dialogue data, including: determining the game analysis result corresponding to each round of dialogue data based on a first preset model; wherein, for the current round of dialogue data, the first preset model is used to obtain the game process information corresponding to the current round of dialogue data based on the first prompt information, and to obtain the game analysis result corresponding to the current round of dialogue data based on the game process information.
[0057] In some optional embodiments, the game process information includes at least one of the following: summary information of the current round of dialogue data, summary information of the other user's historical dialogue data, a first business strategy against the other user, information on the other user's weaknesses, and a second business strategy corresponding to the information on weaknesses; the game analysis results are used to characterize the attitudes of the first user and the second user toward at least some of the target matters in the preset business.
[0058] Among them, summary information is key and important information extracted from dialogue data; first business strategy and second business strategy refer to the processing strategies adopted for preset business; target matters are several important matters in the preset business.
[0059] For example, if the preset service is a telecommunications product sales service, the first user is the customer, and the second user is the salesperson, the corresponding first prompt message may include:
[0060] {
[0061] "Dialogue Round Number": "<The current dialogue round number>",
[0062] Speaker: "Customer or Salesperson"
[0063] "Speech Summary Information": "Based on the above, briefly summarize the main point of the speech in this round."
[0064] "Summary information of the other user's historical dialogue data": "A summary of the historical dialogue data of the other user from the initial conversation to the current round; if the speaker is a customer, then the other user is a salesperson, and vice versa."
[0065] "Opponent user's weakness information": "Identify the opponent's user's weaknesses in order to formulate your own business strategy."
[0066] "The speaker addressed the business strategies adopted by the opposing user in this round": {
[0067] "Defense Strategy": "Deflect the questions, demands, or challenges raised in the previous round of speeches (corresponding to the first business strategy)".
[0068] "Attack Strategy": "Proactively raise new questions, demands, or challenges to help advance the agenda (corresponding to the second business strategy)."
[0069] }
[0070] The corresponding game theory analysis results may include:
[0071] {
[0072] "Are there any disagreements between the two parties regarding the scope of service for the communication products?" <bool>,
[0073] "Are there any disagreements between the two parties regarding the service duration of the communication products?" <bool>,
[0074] "Are there any disputes between the two parties regarding the service fees for communication products?" <bool>,
[0075] }
[0076] As mentioned earlier, after obtaining the game analysis results of each round of dialogue data, the target question category of the first round of dialogue data can be determined.
[0077] In some optional embodiments, the target question category corresponding to each round of first dialogue data is determined based on the game analysis results corresponding to each round of first dialogue data, including: for any round of first dialogue data, selecting one of multiple preset question categories as the target question category of the first dialogue data based on the game analysis results corresponding to the first dialogue data.
[0078] Therefore, for the first user, the possible problems were summarized and analyzed in advance, and corresponding categories were extracted based on these problems, resulting in multiple preset problem categories. Based on this, for any round of first dialogue data currently being processed, the one that best matches it can be selected from the multiple preset problem categories as the target problem category.
[0079] In some optional embodiments, based on the game analysis results corresponding to the first dialogue data, one of multiple preset question categories is selected as the target question category of the first dialogue data, including: selecting multiple candidate question categories that match the first dialogue data from multiple preset question categories; and selecting the one that best matches the game analysis results corresponding to the first dialogue data from multiple candidate question categories as the target question category of the first dialogue data.
[0080] Therefore, for any given round of first dialogue data, the first dialogue data may correspond to multiple of the multiple preset question categories. The multiple preset question categories are the multiple candidate question categories. Based on this, the most matching one can be selected from the multiple candidate question categories as the target question category for the current first dialogue data by combining the game analysis results of the first dialogue data.
[0081] In other words, in any round of the first dialogue data, the first user may ask multiple questions, which are of different importance. Therefore, it is necessary to select the most important question and use the preset question category to which the question belongs as the target question category.
[0082] In some optional embodiments, for any round of first dialogue data, the strategy information adopted by the second user for the first dialogue data can also be determined based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data.
[0083] In some optional embodiments, the strategy information includes a dialogue strategy; correspondingly, based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data, the strategy information adopted by the second user for the first dialogue data is determined, including: determining the second user's strategy action information based on the second dialogue data associated with the first dialogue data; determining the second user's business script information based on the game analysis results corresponding to the associated second dialogue data; and determining the dialogue strategy adopted by the second user for the first dialogue data based on the second user's strategy action information and business script information.
[0084] The second dialogue data associated with the first dialogue data includes the next round of second dialogue data adjacent to the first dialogue data. The strategy action information is a sequence of one or more actions taken by the second user when facing the first user in the target question category corresponding to the current first dialogue data. Each action corresponds to a strategy action, and the corresponding strategy action information can be represented by a text data with a verb-object structure, such as a text description composed of a verb, modifier, and object. The business dialogue information is the response dialogue taken by the second user in response to the first dialogue data. Accordingly, the dialogue strategy can be represented as <strategy action information, business dialogue information>.
[0085] Furthermore, after determining the dialogue strategy, it is also necessary to obtain the problem strategy correspondence result, which includes the target problem category and its corresponding dialogue strategy. In other words, it is necessary to match the target problem category with the dialogue strategy so that the appropriate dialogue strategy can be determined for the problem category to be processed.
[0086] For example, when the second user processes the preset business in the future, he / she can use the result of the question strategy to determine the strategy action and corresponding business script for the first dialogue data of a certain first user after clarifying the target question category, and then determine the corresponding second dialogue data.
[0087] In some alternative embodiments, the relevant data processing in steps S23 to S25 can be implemented using a model.
[0088] For example, the second preset model can obtain the corresponding result of the problem strategy based on the second prompt information.
[0089] For example, the first user is a customer and the second user is an agent. Accordingly, the second prompt information may include: task requirements and output requirements.
[0090] Task requirements:
[0091] 1. The "Target Question Category" should preferably use categories from the "Preset Question Category Library," which includes multiple preset question categories.
[0092] 2. If a dialogue round contains multiple questions, do not merge them. Instead, select the most appropriate, reasonable, and crucial pre-defined question category as the target question category for output.
[0093] 3. An "action strategy" is a sequence of behaviors that a second user takes when faced with a "target problem category," with each behavior corresponding to a strategy action.
[0094] 4. "Business scripts" must correspond one-to-one with strategic actions.
[0095] 5. The "Seat Strategy Summary" is a generalization of the "Motion Strategy."
[0096] Output requirements:
[0097] {
[0098] Current dialogue round: <int>,
[0099] "Analyze the game structure and dialogue content of the current round of dialogue": <str>,
[0100] "Is a preceding text required?" <bool>,
[0101] "Target Problem Category": <str>,
[0102] "Seat strategy": [{"<str, strategy action>": "<str, business script>"}],
[0103] "Summary of seat strategy" <str>
[0104] }
[0105] Figure 3 This is a schematic diagram illustrating the processing procedure of a dialogue data processing method provided in an embodiment of this disclosure. (Refer to...) Figure 3 First, the multi-round dialogue data is input into the first preset model. After processing, the game analysis results corresponding to each round of dialogue data are output. Then, the data is input into the second preset model. After processing by the second preset model, the results corresponding to the problem strategy are output.
[0106] Therefore, the advantage of large models includes extracting the required information from unstructured text data. However, in real-world scenarios, customers' expressions are often very brief and obscure, and agents usually use milder and less direct language when responding. Thus, it is difficult to truly obtain the required content from surface semantics using large models. Based on this, game theory analysis is introduced into the processing. Accordingly, the processing of dialogue data is divided into two parts: the first part mainly involves game theory analysis to obtain the game theory analysis results, and the second part uses the obtained game theory analysis results for in-depth mining to obtain the final problem-strategy correspondence results.
[0107] In summary, introducing game theory analysis into the processing of dialogue data is equivalent to constructing multiple intermediate tasks. Through the output of these intermediate tasks, the original multi-turn dialogue data text is transformed into "information-rich" text, thereby reducing the complexity of task processing and improving the accuracy and rationality of the final output.
[0108] In summary, in this embodiment, game theory processing is introduced into the processing of preset business operations. Game theory analysis is performed on each round of dialogue data to obtain a game theory analysis result corresponding to each round of dialogue data. This game theory analysis result is an intermediate structure in the reasoning process, equivalent to breaking down a large-scale task into multiple intermediate tasks. The corresponding intermediate structure is obtained through processing these intermediate tasks. This processing method can improve the stability of task processing and also allow for sufficient feature extraction and other processing from the initial multi-round dialogue data, thereby obtaining an "information-rich" intermediate structure that can truly reflect the game process and thinking of the two parties in the dialogue; furthermore... Furthermore, based on the game analysis results corresponding to the first dialogue data in each round, the target question category corresponding to each round of the first dialogue data can be determined. Then, for any round of the first dialogue data, based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data, the strategy information adopted by the second user for the first dialogue data can be determined. Finally, based on the question categories of multiple rounds of the first dialogue data and the strategy information corresponding to each round of the first dialogue data, the corresponding result of the question strategy can be obtained. Since an intermediate structure is introduced in the processing, not only can stability be improved, but also the feature extraction effect can be strengthened, thereby improving the accuracy of reasoning.
[0109] It is understood that the various method embodiments mentioned above in this disclosure can be combined with each other to form combined embodiments without violating the principle and logic. Due to space limitations, this disclosure will not elaborate further. Those skilled in the art will understand that in the above methods of specific implementation, the specific execution order of each step should be determined by its function and possible internal logic.
[0110] Secondly, embodiments of this disclosure provide a dialogue data processing apparatus.
[0111] Figure 4 This is a block diagram of a dialogue data processing apparatus provided in an embodiment of the present disclosure.
[0112] Reference Figure 4 This disclosure provides a dialogue data processing apparatus 400, which may include the following modules.
[0113] The first acquisition module 401 is used to acquire multi-round dialogue data between the first user and the second user for processing preset business. The multi-round dialogue data is composed of alternating first dialogue data and second dialogue data, with the first dialogue data corresponding to the first user and the second dialogue data corresponding to the second user.
[0114] Analysis module 402 is used to perform game analysis on the dialogue data of each round and obtain the game analysis results corresponding to the dialogue data of each round.
[0115] The first determining module 403 is used to determine the target question category corresponding to each round of the first dialogue data based on the game analysis results corresponding to the first dialogue data in each round.
[0116] The second determining module 404 is used to determine the strategy information adopted by the second user for the first dialogue data based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data for any round of first dialogue data.
[0117] The second acquisition module 405 is used to obtain the result corresponding to the question policy based on the question category of the multi-round first dialogue data and the policy information corresponding to each round of first dialogue data.
[0118] The embodiments provided in this disclosure introduce game theory processing into the processing of preset business operations. Game theory analysis is performed on each round of dialogue data to obtain a game theory analysis result corresponding to each round of dialogue data. This game theory analysis result is an intermediate structure in the reasoning process, which is equivalent to breaking down a large-scale task into multiple intermediate tasks. The corresponding intermediate structure is obtained through the processing of intermediate tasks. This processing method can improve the stability of task processing on the one hand, and fully extract features from the initial multi-round dialogue data on the other hand, thereby obtaining an "information-rich" intermediate structure that can truly reflect the game process and game thinking of the two parties in the dialogue. Furthermore, the target question category corresponding to each round of first dialogue data can be determined based on the game theory analysis result corresponding to each round of first dialogue data. Then, for any round of first dialogue data, the strategy information adopted by the second user for the first dialogue data can be determined based on the second dialogue data associated with the first dialogue data and the game theory analysis result corresponding to the associated second dialogue data. Finally, the question strategy corresponding result can be obtained based on the question category of multi-round first dialogue data and the strategy information corresponding to each round of first dialogue data. Since an intermediate structure is introduced in the processing, not only can stability be improved, but the feature extraction effect can also be enhanced, thereby improving the accuracy of reasoning.
[0119] Each module in the aforementioned dialogue data processing device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the operations corresponding to each module.
[0120] In addition, this disclosure also provides electronic devices and computer-readable storage media, all of which can be used to implement any of the dialogue data processing methods provided in this disclosure. The corresponding technical solutions and descriptions are described in the corresponding descriptions in the method section, and will not be repeated here.
[0121] Figure 5 This is a block diagram of an electronic device provided in an embodiment of the present disclosure.
[0122] Reference Figure 5 This disclosure provides an electronic device, which includes: at least one processor 501 connected via a bus; at least one memory 502; and one or more I / O interfaces 503; wherein the memory 502 stores one or more computer programs that can be executed by the at least one processor 501, and the one or more computer programs are executed by the at least one processor 501 to enable the at least one processor 501 to perform the above-described dialogue data processing method.
[0123] The modules in the aforementioned electronic devices can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.
[0124] This disclosure also provides a computer-readable storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the above-described dialogue data processing method. The computer-readable storage medium may be volatile or non-volatile.
[0125] This disclosure also provides a computer program product, including computer-readable code, or a non-volatile computer-readable storage medium carrying computer-readable code, wherein when the computer-readable code is run in a processor of an electronic device, the processor in the electronic device executes the above-described dialogue data processing method.
[0126] Those skilled in the art will understand that all or some of the steps, systems, and apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software can be distributed on a computer-readable storage medium, which may include computer storage media (or non-transitory media) and communication media (or transient media).
[0127] As is known to those skilled in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable program instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), static random access memory (SRAM), flash memory or other memory technologies, portable compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer. Furthermore, it is known to those skilled in the art that communication media typically contain computer-readable program instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.
[0128] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.
[0129] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages such as Smalltalk, C++, etc., and conventional procedural programming languages such as "C" or similar languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.
[0130] The computer program product described herein can be implemented specifically through hardware, software, or a combination thereof. In one alternative embodiment, the computer program product is specifically embodied in a computer storage medium; in another alternative embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.
[0131] Various aspects of this disclosure are described herein with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer-readable program instructions.
[0132] These computer-readable program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that, when executed by the processor of the computer or other programmable data processing apparatus, they create means for implementing the functions / actions specified in one or more blocks of the flowchart and / or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium that causes a computer, programmable data processing apparatus, and / or other device to operate in a particular manner; thus, the computer-readable medium storing the instructions comprises an article of manufacture that includes instructions for implementing aspects of the functions / actions specified in one or more blocks of the flowchart and / or block diagram.
[0133] Computer-readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other device to produce a computer-implemented process, thereby causing the instructions executed on the computer, other programmable data processing apparatus, or other device to perform the functions / actions specified in one or more boxes of a flowchart and / or block diagram.
[0134] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of an instruction containing one or more executable instructions for implementing a specified logical function. In some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions.
[0135] Example embodiments have been disclosed herein, and while specific terminology has been used, it is for illustrative purposes only and should be construed as such, and is not intended to be limiting. In some instances, it will be apparent to those skilled in the art that features, characteristics, and / or elements described in connection with particular embodiments may be used alone, or in combination with features, characteristics, and / or elements described in connection with other embodiments, unless otherwise expressly indicated. Therefore, those skilled in the art will understand that various changes in form and detail may be made without departing from the scope of this disclosure as set forth by the appended claims.< / str> < / str> < / bool> < / str> < / int> < / bool> < / bool> < / bool>
Claims
1. A dialogue data processing method, characterized in that, The method includes: Acquire multi-round dialogue data between a first user and a second user for processing preset business. The multi-round dialogue data consists of alternating first dialogue data and second dialogue data, where the first dialogue data corresponds to the first user and the second dialogue data corresponds to the second user. Game analysis is performed on the dialogue data for each round to obtain the game analysis results corresponding to each round of dialogue data; Based on the game analysis results corresponding to the first dialogue data in each round, determine the target question category corresponding to the first dialogue data in each round. For any round of first dialogue data, based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data, determine the strategy information adopted by the second user for the first dialogue data; Based on the question categories of the first dialogue data from multiple rounds and the policy information corresponding to the first dialogue data in each round, the corresponding result of the question policy is obtained.
2. The method according to claim 1, characterized in that, The step of performing game analysis on the dialogue data for each round to obtain the game analysis results corresponding to each round of dialogue data includes: Perform game theory analysis on each round of dialogue data to obtain the game theory analysis results corresponding to each round of dialogue data; or, For the i-th round of dialogue data, perform game analysis based on the i-1 game analysis results corresponding to the i-th round of dialogue data and the previous i-1 rounds of dialogue data, and obtain the game analysis result corresponding to the i-th round of dialogue data, where i>1.
3. The method according to claim 2, characterized in that, The process of performing game analysis on each round of dialogue data to obtain game analysis results corresponding to each round of dialogue data includes: The game analysis results corresponding to each round of dialogue data are determined based on the first preset model; Specifically, for the current round of dialogue data, the first preset model is used to obtain the game process information corresponding to the current round of dialogue data based on the first prompt information, and to obtain the game analysis result corresponding to the current round of dialogue data based on the game process information.
4. The method according to claim 3, characterized in that, The game process information includes at least one of the following: summary information of the current round of dialogue data, summary information of the opponent user's historical dialogue data, a first business strategy targeting the opponent user, information on the opponent user's weaknesses, and a second business strategy corresponding to the information on the weaknesses. The game theory analysis results are used to characterize the attitudes of the first user and the second user towards at least some of the target matters in the preset business.
5. The method according to claim 1, characterized in that, The step of determining the target question category corresponding to the first dialogue data in each round based on the game analysis results includes: For any given round of first dialogue data, based on the game analysis results corresponding to the first dialogue data, one of multiple preset question categories is selected as the target question category for the first dialogue data.
6. The method according to claim 5, characterized in that, The step of selecting one of multiple preset question categories as the target question category for the first dialogue data based on the game analysis results corresponding to the first dialogue data includes: Select from multiple preset question categories that match the first dialogue data; Select the one that best matches the game analysis result corresponding to the first dialogue data from among the multiple candidate question categories as the target question category of the first dialogue data.
7. The method according to claim 1, characterized in that, The strategy information includes dialogue strategies; The step of determining the strategy information adopted by the second user in relation to the first dialogue data based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data includes: Based on the second dialogue data associated with the first dialogue data, determine the policy action information of the second user; Based on the game analysis results corresponding to the associated second dialogue data, determine the business script information of the second user; Based on the second user's policy action information and business script information, determine the dialogue strategy adopted by the second user in response to the first dialogue data.
8. A dialogue data processing device, characterized in that, The device includes: The first acquisition module is used to acquire multi-round dialogue data between a first user and a second user for processing preset business. The multi-round dialogue data is composed of alternating first dialogue data and second dialogue data, where the first dialogue data corresponds to the first user and the second dialogue data corresponds to the second user. The analysis module is used to perform game analysis on the dialogue data of each round and obtain the game analysis results corresponding to each round of dialogue data. The first determining module is used to determine the target question category corresponding to the first dialogue data in each round based on the game analysis results corresponding to the first dialogue data in each round. The second determining module is used to determine the strategy information adopted by the second user for the first dialogue data based on the second dialogue data associated with the first dialogue data and the game analysis results corresponding to the associated second dialogue data for any round of first dialogue data. The second acquisition module is used to obtain the result corresponding to the question policy based on the question category of the first dialogue data in multiple rounds and the policy information corresponding to the first dialogue data in each round.
9. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores one or more computer programs that can be executed by the at least one processor, the one or more computer programs being executed by the at least one processor to enable the at least one processor to perform the dialogue data processing method as described in any one of claims 1-7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the dialogue data processing method as described in any one of claims 1-7.