Dialogue information processing method and device, equipment and medium

Abstract

The invention relates to a dialogue information processing method and device, equipment and a medium, and relates to the technical field of artificial intelligence, and the method comprises the steps: determining a dialogue user; performing user information extraction on the dialogue user to obtain user information of the dialogue user; and performing chat processing based on the user information in combination with the current dialogue information of the dialogue user to obtain chat response information of the dialogue user, thereby realizing long-term memory in multiple rounds of dialogues. The problems that in the prior art, historical information is lost, and long-term memory of the user cannot be achieved in a multi-round dialogue scene are solved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence (AI) technology, and in particular to a method, device, equipment and medium for processing dialogue information. Background Art

[0002] With the rapid development of AI technology, multi-round dialogue models implemented using AI technology are widely used in multi-round dialogue chat scenarios in various industries, such as online customer service, smart speakers, AI chat robots, social media automatic reply systems, etc. Multi-round dialogue models are used to achieve long-term dialogues with users to improve the user experience in long-term dialogues.

[0003] Currently, the existing multi-round dialogue model is limited by the length of the model input and can often only remember the most recent few rounds of dialogue. There is no effective user information update and storage mechanism. As the number of dialogue rounds increases, early dialogue information is easily forgotten, resulting in the forgetting of important historical information such as user personal information, preferences, and interaction events in historical dialogues during the multi-round dialogue process, which affects the effect of interaction with users. Summary of the invention

[0004] The present application provides a conversation information processing method, device, equipment and medium to solve the problem of historical information loss caused by the increase in the number of conversation rounds in the existing related technology in a multi-round conversation chat scenario, thereby improving the effect of interaction with users.

[0005] In a first aspect, the present application provides a method for processing dialogue information, comprising:

[0006] Identify the conversation user;

[0007] Extracting user information from the conversation rounds to obtain user information of the conversation users;

[0008] Based on the user information, chat processing is performed in combination with the current conversation information of the conversation user to obtain chat response information of the conversation user.

[0009] Optionally, extracting user information of the conversation user to obtain the user information of the conversation user includes:

[0010] Determining the number of conversation turns of the conversation users;

[0011] When the number of conversation rounds reaches a preset extraction round number threshold, obtaining the historical conversations of the conversation user;

[0012] User information is extracted based on the historical conversation to obtain user information of the conversation user.

[0013] Optionally, extracting user information based on the historical conversation to obtain user information of the conversation user includes:

[0014] Extracting unstructured information related to the conversation users from the historical conversation;

[0015] Based on the unstructured information, the original user information of the conversation user is updated to obtain the remembered user information;

[0016] The memorized user information is determined as the user information of the conversation user.

[0017] Optionally, the unstructured information includes personal preference information of the conversation user and interaction event information in the historical conversation, and the updating of original user information of the conversation user based on the unstructured information to obtain the remembered user information includes:

[0018] extracting new user information from the unstructured information based on the personal preference information and the interaction event information;

[0019] Merging the new user information with the original user information to obtain merged user information;

[0020] The new user information is used as the basis, and the conflicting parts in the merged user information are removed to obtain the remembered user information.

[0021] Optionally, based on the user information, chat processing is performed in combination with the current conversation information of the conversation user to obtain the chat response information of the conversation user, including:

[0022] Determining a chat model used by the conversation users;

[0023] Inputting the user information and the current conversation information into the chat model;

[0024] The chat processing is performed through the chat model to generate the chat response information.

[0025] Optionally, before inputting the user information and the current dialogue information into the chat model, the process further includes: obtaining question information input by the dialogue user; and determining the current dialogue information using the question information.

[0026] Optionally, determining the number of conversation turns of the conversation users includes:

[0027] After the dialog user logs in, based on the question information input by the dialog user, the number of dialog inputs corresponding to the dialog user is recorded;

[0028] The number of dialogue turns is determined by using the number of dialogue inputs.

[0029] In a second aspect, the present application provides a dialogue information processing device, comprising:

[0030] A dialogue user determination module, used to determine the dialogue user;

[0031] A user information extraction module, used to extract user information of the conversation user to obtain user information of the conversation user;

[0032] The chat processing module is used to perform chat processing based on the user information and the current conversation information of the conversation user to obtain the chat response information of the conversation user.

[0033] In a third aspect, the present application provides an electronic device comprising: a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; the memory is used to store computer programs; the processor is used to implement the steps of the dialogue information processing method as described in any one of the above-mentioned first aspects when executing the program stored in the memory.

[0034] In a fourth aspect, the present application further provides a computer storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute the steps of the dialogue information processing method described in any one of the first aspects above.

[0035] The conversation information processing method, apparatus, device and medium provided in the embodiments of the present application determine the conversation user and extract user information of the conversation user, perform chat processing based on the user information of the conversation user and the current conversation information of the conversation user, and obtain the chat response information of the conversation user, thereby realizing long-term memory of multiple rounds of conversations and avoiding the loss of historical information. It also solves the problems of historical information loss and the inability to realize long-term memory of users in multiple rounds of conversation scenarios existing in the existing related technologies, thereby improving the user experience in long-term conversations. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0038] One or more embodiments are exemplarily described by pictures in the corresponding drawings, and these exemplified descriptions do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings represent similar elements, and unless otherwise stated, the figures in the drawings do not constitute proportional limitations.

[0039] Figure 1 A schematic diagram of a process flow of a method for processing a conversation message provided in an embodiment of the present application;

[0040] Figure 2 A schematic diagram of a multi-round conversation chat process provided as an example of this application;

[0041] Figure 3 A schematic diagram of implementing multi-round conversation chat processing by extracting user information provided in an embodiment of the present application;

[0042] Figure 4 A structural block diagram of a conversation information processing device provided in an embodiment of the present application;

[0043] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0044] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0045] The disclosure below provides many different embodiments or examples to implement different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or letters in different examples. This repetition is for the purpose of simplification and clarity, and does not itself indicate the relationship between the various embodiments and / or settings discussed.

[0046] In the existing related technologies, user information extraction usually uses Named Entity Recognition (NER) technology to identify specific nouns mentioned in the user input information value, such as identifying specific types of structured information such as names of people, places, and organization names. However, other unstructured information cannot be effectively extracted and remembered. For example, it is impossible to extract and remember unstructured information such as user preferences and emotions. In addition, the current multi-round dialogue model has no effective mechanism to update and store this user information, which leads to the forgetting of early user information such as user personal information, preferences, and interaction events in historical dialogues during the multi-round dialogue process. There are problems of historical information loss and inability to achieve long-term memory of users in multi-round dialogue scenarios.

[0047] Based on the above, the present application provides a method for processing conversation information, by determining the conversation users and extracting user information based on the conversation users, so as to perform chat processing based on the user information of the conversation users and the current conversation information of the conversation users, so as to realize long-term memory of user information in multiple rounds of conversations, that is, to realize long-term memory of multiple rounds of conversations, avoid the loss of historical information, solve the problems of historical information loss and the inability to realize long-term memory of users in multiple rounds of conversation scenarios existing in the existing related technologies, and can effectively improve the user experience in long-term conversations.

[0048] Figure 1 A flowchart of a method for processing dialog information provided by an embodiment of the present application. Figure 1 As shown, the dialogue information processing method provided in the embodiment of the present application may specifically include the following steps:

[0049] Step 110, determining the dialogue user;

[0050] Among them, conversation users refer to users participating in chat conversations, which may specifically include users who conduct multiple rounds of conversations using a round-by-round conversation model, such as users who conduct conversations with online customer service, smart speakers, AI chat robots, or social media automatic reply systems. This embodiment of the present application does not limit this.

[0051] Step 120, extracting user information based on the number of conversation rounds to obtain user information of the conversation user;

[0052] Specifically, after determining the conversation user, the embodiment of the present application can extract user information for the conversation user. For example, based on the number of conversation turns of the conversation user, the characteristics of the large language model (LLM) can be used to regularly extract non-structural information related to the conversation user from historical conversations, and merge the information with the original user information of the conversation user, remove conflicting parts, and update the newly added parts to obtain new user information as the user information of the conversation user, so that subsequent chat processing can be performed based on the latest user information of the conversation user, that is, execute step 130, thereby assisting in generating chat responses that are more inclined to the user.

[0053] It should be noted that the conversation round number of the dialogue user may refer to the conversation round number of the dialogue information currently input by the dialogue user. For example, in the scenario of using AI technology to realize multi-round dialogue chat, when the dialogue user inputs "hello" as the question information of the first round of dialogue, Figure 2 As shown, when the question information input by the user is "Hello", it can be determined that the number of conversation turns of the dialogue user is 1; and when the dialogue user inputs "My name is Xiaodou" as the question information of the second round of dialogue, that is, when the question information input by the user is "My name is Xiaodou", it is determined that the number of conversation turns of the dialogue user is 2... and so on, when the dialogue user inputs "Of course it's good, photography is one of my specialties, and one of my wishes in life is that the photos I take can be selected into the National Geographic Magazine" as the question information of the fifth round of dialogue, that is, when the question information input by the user is "Of course it's good, photography is one of my specialties, and one of my wishes in life is that the photos I take can be selected into the National Geographic Magazine", it is determined that the number of conversation turns of the dialogue user is 5.

[0054] Step 130: Perform chat processing based on the user information and in combination with the current conversation information of the conversation user to obtain chat response information of the conversation user.

[0055] Among them, the current conversation information of the conversation user may refer to the conversation information currently obtained for the conversation user, such as the conversation information currently input by the conversation user may be obtained as the current conversation information, and the embodiment of the present application does not limit this. Specifically, in this step, the embodiment of the present application may use the user information obtained in step 120 as an additional input source to use the user information to perform chat processing on the conversation information currently input by the conversation user, thereby assisting in generating a chat response that is more inclined to the conversation user, as the chat response information of the conversation user, and outputting it to the conversation user, thereby effectively improving the user chat experience in the multi-round conversation scenario.

[0056] It can be seen that the embodiment of the present application determines the dialogue user and extracts user information of the dialogue user to obtain the user information of the dialogue user, and then performs chat processing based on the user information in combination with the current dialogue information of the dialogue user, so as to use the user information to assist in generating a chat response that is more inclined to the user, and obtain the chat response information of the dialogue user. This can effectively avoid the problem of historical information loss caused by an increase in the number of dialogue rounds, thereby ensuring the accuracy of chat response information generation, improving the user chat experience in multi-round dialogue scenarios, and solving the problem of historical information loss caused by an increase in the number of dialogue rounds in the existing related technologies in multi-round dialogue chat scenarios, thereby improving the effect of interaction with users.

[0057] In a specific implementation, the dialogue information processing method provided in the embodiment of the present application can be widely used in turn-to-turn dialogue chat scenarios, such as long-term dialogue scenarios with AI intelligent software such as online customer service, smart speakers, AI chat robots, and social media automatic reply systems. By determining the number of dialogue rounds of the dialogue users, user information can be extracted after a certain number of rounds of chat, so that the extracted user information can be used for subsequent chat processing.

[0058] In some optional embodiments of the present application, the number of conversation rounds of the conversation user can be determined, so that when the number of conversation rounds reaches a preset extraction round threshold, the historical conversation of the conversation user can be obtained, so that user information can be extracted based on the historical conversation to obtain the user information of the conversation user.

[0059] In an embodiment of the present application, the conversation round number of the conversation user can be determined by obtaining the question information input by the conversation user, so as to extract the user information of the conversation user based on the conversation round number to obtain the user information of the conversation user. Optionally, the embodiment of the present application determines the conversation round number of the conversation user, which can specifically include: after the conversation user logs in, based on the question information input by the conversation user, recording the number of conversation inputs corresponding to the conversation user; and using the number of conversation inputs to determine the number of conversation rounds.

[0060] The number of dialogue inputs may refer to the number of times a dialogue user inputs information in the same dialogue process. For example, in a long dialogue process, the dialogue user first inputs "hello" as the question information of the first round of dialogue. Figure 2 As shown, the number of dialogue inputs recorded at this time is 1, and the number of dialogue inputs recorded when the user enters "My name is Xiaodou" as the question information for the second round of dialogue is 2... and so on. When the user enters "Of course it's good, photography is one of my specialties, and one of my life wishes is that the photos I take can be selected into National Geographic Magazine" as the question information for the fifth round of dialogue, the number of dialogue inputs recorded is 5.

[0061] After determining the number of conversation rounds of the conversation user, the embodiment of the present application can determine whether it is necessary to extract user information by judging whether the number of conversation rounds reaches a preset extraction round threshold, so as to solve the problem of information loss caused by the increase of conversation rounds in the existing related technology in the long-term conversation scenario. Optionally, the embodiment of the present application extracts user information based on the number of conversation rounds to obtain the user information of the conversation user, which can specifically include the following sub-steps:

[0062] Sub-step S11, when the number of conversation rounds reaches a preset extraction round number threshold, obtaining the historical conversations of the conversation user;

[0063] Sub-step S12, extracting user information based on the historical conversation to obtain user information of the conversation user.

[0064] Among them, the extraction round threshold can be set according to the needs of user information extraction in multi-round dialogue chats, so that user information can be extracted regularly according to a certain number of dialogue rounds. For example, user information can be extracted every few rounds of dialogue to solve the problem of information loss caused by the increase in the number of dialogue rounds in multi-round dialogue chat scenarios.

[0065] Specifically, the conversation information processing method provided by the implementation of the present application is applied to a multi-round conversation chat scenario. After determining the number of conversation rounds of the conversation user, the conversation round number can be compared with a preset extraction round number threshold to determine whether the conversation user's current number of conversation rounds has reached the preset extraction round number threshold; if the user's current number of conversation rounds has reached the preset extraction round number threshold, the conversation user's historical conversation can be obtained, and user information can be extracted based on the historical conversation, so that user information can be extracted through cyclic extraction after the chat reaches a certain number of conversation rounds, thereby realizing regular extraction of user information, thereby realizing long-term memory of user information in multi-round conversation scenarios, and solving the problem of information loss caused by increasing the number of conversation rounds.

[0066] In some optional embodiments of the present application, user information is extracted based on the historical conversation to obtain the user information of the conversation user, which may specifically include: extracting unstructured information related to the conversation user from the historical conversation; updating the original user information of the conversation user based on the unstructured information to obtain the remembered user information; and determining the remembered user information as the user information of the conversation user. Among them, unstructured information refers to unstructured information, which may specifically include unstructured information that cannot be extracted using named entity recognition technology, such as the user's personal preference information, interactive events in historical conversations, etc.; the original user information of the conversation user may refer to the original user information of the conversation user, such as the original user information may include the name, gender, age, occupation, personal preferences, etc. of the conversation user, and the embodiments of the present application do not impose specific restrictions on this.

[0067] Optionally, when the unstructured information related to the conversation user includes the personal preference information of the conversation user and / or the interaction event information in the historical conversation, the embodiment of the present application updates the original user information of the conversation user based on the unstructured information to obtain the remembered user information, which may specifically include: extracting new user information from the unstructured information based on the personal preference information and the interaction event information; merging the new user information with the original user information to obtain merged user information; and removing the contradictory parts in the merged user information based on the new user information to obtain the remembered user information.

[0068] Among them, personal preference information can be used to determine the user's personal and preference information, and specifically may include the user's personal information and preference information, such as name, gender, age, occupation, personal preferences and other information, and the embodiments of the present application do not impose specific restrictions on this; interaction event information may include various information recorded for interaction events in historical conversations, and specifically may be used to determine interaction events in historical conversations.

[0069] Merged user information refers to the new user information generated by merging the extracted new user information with the original user information. For example, when the original user information of the conversation user is recorded as "[Name] Xiaodou; [Gender] Male; [Age] 20; [Occupation] High school student; [Personal preference] Food", if the new user information extracted from the unstructured information is "[Name] Xiaodou; [Gender] Male; [Age] 20; [Occupation] College student; [School] XX University; [Personal specialty] Photography; [Personal preference] I hope that the photos I take can be selected into the National Geographic Magazine", then by merging the new user information with the original user information, removing the contradictory parts, and updating the newly added parts, the memory user information obtained is "[Name] Xiaodou; [Gender] Male; [Age] 20; [Occupation] College student; [School] XX University; [Personal specialty] Photography; [Personal preference] I hope that the photos I take can be selected into the National Geographic Magazine, Food".

[0070] It can be seen that the embodiment of the present application extracts unstructured information related to the user from the historical conversations, and can update the original user information of the conversation user based on the extracted unstructured information, so that the updated user information can be used as the user information of the conversation user, so that the updated user information can be used to perform chat processing on the currently acquired conversation information later, which not only solves the problem of information loss caused by the increase in the number of conversation rounds in the prior art, but also solves the problem of information contradiction, and can achieve long-term memory of user information under limited historical conversation input conditions.

[0071] In some optional embodiments of the present application, the characteristics of a large language model (LLM) can be utilized to extract and update various user information related to the conversation users, so that the chat model can achieve long-term memory under limited historical conversation input conditions, thereby improving the user experience in long-term conversations. Among them, the chat model may include a multi-round conversation model applied to multi-round conversations. For example, a large language model LLM can be used as the chat model in the embodiments of the present application, and the embodiments of the present application are not limited to this. Optionally, the embodiments of the present application perform chat processing based on the user information and the current conversation information of the conversation user to obtain the chat response information of the conversation user, which may specifically include the following sub-steps:

[0072] Sub-step S21, determining the chat model used by the dialogue user;

[0073] Sub-step S22, inputting the user information and the current conversation information into the chat model;

[0074] Sub-step S23, performing chat processing through the chat model to generate the chat response information.

[0075] Specifically, the embodiment of the present application determines the number of conversation rounds of the conversation users, and extracts user information when the number of conversation rounds reaches a preset extraction round threshold, and can clear the recorded number of conversation inputs to solve the problem of information loss caused by the increase in the number of conversation rounds by regularly extracting user information.

[0076] In addition, after the user information is extracted, the embodiment of the present application inputs the user information of the conversation user into the chat model as an additional input source for the chat model, so that the chat model can perform chat processing on the current conversation information of the conversation user based on the user information. For example, the chat model can perform chat processing on the question information currently input by the conversation user based on the user data, so as to utilize the user information to assist the chat model in generating a chat response that is more inclined to the user, which is output as the chat response information of the conversation user, thereby improving the user experience in long-term conversations.

[0077] It should be noted that the question information in the embodiment of the present application may include information input by the dialogue users during multiple rounds of dialogue and chat, such as the information "Hello" input by the user during the first round of dialogue.

[0078] Optionally, based on the above embodiment, the conversation information processing method provided by the present application may further include the following steps before inputting the user information and the current conversation information into the chat model:

[0079] Step 210, obtaining question information input by the dialog user;

[0080] Step 220: Using the question information to determine the current dialogue information.

[0081] Specifically, after obtaining the question information input by the dialogue user, the embodiment of the present application can input the question information into the chat model as the current dialogue information of the dialogue user, so that the chat model can perform chat processing on the question information based on the user information of the dialogue user, generate and output corresponding chat response information, and form a new dialogue with the dialogue user, and then automatically realize multiple rounds of dialogue with the dialogue user, thereby improving the user experience.

[0082] As an example of the present application, after a certain number of chat rounds, user information can be extracted regularly. For example, when the preset extraction round threshold is 10, unstructured information related to the conversation user, such as name, gender, age, date of birth, constellation, zodiac sign, current residence, hometown, education, school, occupation, work experience, emotional state, personality (MBTI), habits, personal preferences, personal expertise, travel records, etc., can be extracted from historical conversations every 10 rounds of conversations, and the original user information can be updated based on the unstructured information, thereby realizing the extraction and update of user information, that is, based on historical conversations and original user information, user information is extracted and updated, such as Figure 3 As shown, the new user information generated after the update can be used as the memory user information and input into the chat model, so as to assist the chat model in generating chat response information that is more inclined to the user by using the user information of the dialogue user as an additional input source of the chat model, thereby ensuring the accuracy of the chat response information and improving the user experience. Specifically, after obtaining the question information input by the dialogue user in the current dialogue, the chat model can perform chat processing on the question information based on the new user information, generate the chat response information corresponding to the question information, and output it to form a new dialogue, so that the chat model can achieve long-term memory under the condition of limited historical dialogue input, and effectively improve the user experience in long-term dialogue.

[0083] It should be noted that, for the purpose of simple description, the method embodiments are expressed as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present application are not limited to the described order of actions, because according to the embodiments of the present application, certain steps may be performed in other orders or simultaneously.

[0084] like Figure 4 As shown, the embodiment of the present application provides a dialogue information processing device, including:

[0085] A dialogue user determination module 410, used to determine a dialogue user;

[0086] A user information extraction module 420, configured to extract user information of the conversation user to obtain user information of the conversation user;

[0087] The chat processing module 430 is used to perform chat processing based on the user information and the current conversation information of the conversation user to obtain chat response information of the conversation user.

[0088] Optionally, the user information extraction module 420 includes:

[0089] A conversation turn determination submodule, used to determine the conversation turns of the conversation users;

[0090] An acquisition submodule, configured to acquire the historical conversations of the conversation user when the number of conversation rounds reaches a preset extraction round number threshold;

[0091] The extraction submodule is used to extract user information based on the historical conversation to obtain the user information of the conversation user.

[0092] Optionally, the extraction submodule includes:

[0093] An extraction unit, configured to extract unstructured information related to the conversation user from the historical conversation;

[0094] An updating unit, configured to update the original user information of the conversation user based on the unstructured information to obtain the remembered user information;

[0095] A determination unit is used to determine the remembered user information as the user information of the conversation user.

[0096] Optionally, the unstructured information includes personal preference information of the conversation user and interaction event information in the historical conversation, and the updating unit is specifically used to: extract new user information from the unstructured information based on the personal preference information and the interaction event information; merge the new user information with the original user information to obtain merged user information; and remove conflicting parts in the merged user information based on the new user information to obtain the remembered user information.

[0097] Optionally, the chat processing module 410 may include the following submodules:

[0098] A model determination submodule, used to determine the chat model used by the dialogue user;

[0099] An information input submodule, used for inputting the user information and the current dialogue information into the chat model;

[0100] The chat processing submodule is used to perform chat processing through the chat model and generate the chat response information.

[0101] Optionally, the dialogue information processing device further includes: a question information module; the question information module is used to: obtain question information input by the dialogue user; and use the question information to determine the current dialogue information.

[0102] For example, before the information input submodule inputs the user information and the current conversation information into the chat model, the question information module obtains the question information input by the conversation user and uses the question information to determine the current conversation information, so that the information input submodule can input the current conversation information and the user information of the conversation user into the chat model for chat processing, so that the chat model can automatically generate chat response information to improve the user experience.

[0103] Optionally, the dialogue turn number determination submodule includes:

[0104] A recording unit, configured to record the number of dialogue inputs corresponding to the dialogue user based on the question information input by the dialogue user after the dialogue user logs in;

[0105] A dialogue round number determination unit is used to determine the dialogue round number using the dialogue input times.

[0106] In a specific implementation, the above-mentioned conversation information processing device can be applied in an electronic device, so that the electronic device acts as a conversation information processing device, determines the conversation user, extracts the user information of the conversation user, and then performs chat processing based on the user information of the conversation user and the current conversation information of the conversation user, so as to realize long-term memory of user information in multiple rounds of conversations, obtain the chat response information of the conversation user, realize long-term memory of multiple rounds of conversations, and provide an effective user information update and storage mechanism for the model applied in the multiple rounds of chat scenarios, solve the problems of historical information loss and inability to realize long-term memory of users in multiple rounds of conversation scenarios in the existing related technologies, and improve the user experience in long-term conversations. The conversation information processing device can be composed of two or more physical entities, or it can be composed of one physical entity, such as the conversation information processing device can be a personal computer (PC), a computer, a server, etc., and the embodiments of the present application do not make specific restrictions on this.

[0107] like Figure 5 As shown, an embodiment of the present application provides an electronic device, including a processor 111, a communication interface 112, a memory 113 and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 communicate with each other through the communication bus 114, and the memory 113 is used to store computer programs; the processor 111 is used to implement the dialogue information processing method provided by any of the aforementioned method embodiments when executing the program stored in the memory 113.

[0108] In one embodiment of the present application, the processor 111 is used to execute the program stored in the memory 113 to implement the conversation information processing method provided by any one of the aforementioned method embodiments, including: determining the conversation user; extracting user information of the conversation user to obtain the user information of the conversation user; based on the user information, combined with the current conversation information of the conversation user, performing chat processing to obtain the chat response information of the conversation user.

[0109] An embodiment of the present application further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the dialogue information processing method provided in any of the aforementioned method embodiments are implemented.

[0110] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0111] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a general hardware platform, and of course, by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the processing method described in each embodiment or some parts of the embodiment.

[0112] It should be understood that the terms used herein are only for the purpose of describing specific example embodiments and are not intended to be limiting. Unless the context clearly indicates otherwise, the singular forms "one", "an" and "said" as used herein may also be meant to include plural forms. The terms "include", "comprise", "contain", and "have" are inclusive, and therefore specify the existence of stated features, steps, operations, elements and / or parts, but do not exclude the existence or addition of one or more other features, steps, operations, elements, parts, and / or combinations thereof. The method steps, processes, and operations described herein are not interpreted as necessarily requiring them to be performed in the specific order described or illustrated, unless the execution order is clearly indicated. It should also be understood that additional or alternative steps may be used.

[0113] The foregoing is merely a specific embodiment of the present invention, which enables those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features claimed herein.

Claims

1. A method for processing dialogue information, characterized in that: include: Identify the conversation user; Extracting user information of the conversation user to obtain user information of the conversation user; Based on the user information, chat processing is performed in combination with the current conversation information of the conversation user to obtain chat response information of the conversation user.

2. The processing method according to claim 1, characterized in that: The extracting user information of the conversation user to obtain the user information of the conversation user includes: Determining the number of conversation turns of the conversation users; When the number of conversation rounds reaches a preset extraction round number threshold, obtaining the historical conversations of the conversation user; User information is extracted based on the historical conversation to obtain user information of the conversation user.

3. The processing method according to claim 1, characterized in that: The extracting user information based on the historical conversation to obtain user information of the conversation user includes: Extracting unstructured information related to the conversation users from the historical conversation; Based on the unstructured information, the original user information of the conversation user is updated to obtain the remembered user information; The memorized user information is determined as the user information of the conversation user.

4. The processing method according to claim 3, characterized in that: The unstructured information includes the personal preference information of the dialogue user and the interaction event information in the historical dialogue. The original user information of the dialogue user is updated based on the unstructured information to obtain the remembered user information, including: extracting new user information from the unstructured information based on the personal preference information and the interaction event information; Merging the new user information with the original user information to obtain merged user information; The new user information is used as the basis, and the conflicting parts in the merged user information are removed to obtain the remembered user information.

5. The processing method according to claim 1, characterized in that: The step of performing chat processing based on the user information and combining the current conversation information of the conversation user to obtain the chat response information of the conversation user includes: Determining a chat model used by the conversation user; Inputting the user information and the current conversation information into the chat model; The chat processing is performed through the chat model to generate the chat response information.

6. The processing method according to claim 5, characterized in that: Before inputting the user information and the current dialogue information into the chat model, the method further includes: Obtain question information input by the dialog user; The current dialog information is determined using the question information.

7. The processing method according to any one of claims 2 to 6, characterized in that: The determining of the number of dialogue turns of the dialogue users includes: After the dialog user logs in, based on the question information input by the dialog user, the number of dialog inputs corresponding to the dialog user is recorded; The number of dialogue turns is determined by using the number of dialogue inputs.

8. A conversation information processing device, characterized in that: include: A dialogue user determination module, used to determine the dialogue user; A user information extraction module, used to extract user information of the conversation user to obtain the user information of the conversation user; The chat processing module is used to perform chat processing based on the user information and the current conversation information of the conversation user to obtain the chat response information of the conversation user.

9. An electronic device, characterized in that: include: A processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus; Memory, used to store computer programs; The processor is used to implement the steps of the dialogue information processing method as described in any one of claims 1 to 7 when executing the program stored in the memory.

10. A computer storage medium storing computer executable instructions, wherein the computer executable instructions are used to execute the steps of the dialogue information processing method according to any one of claims 1 to 7.

Images