Question-answering methods, devices, equipment, and media used in live streaming

By generating response information in the live broadcast room through a multi-agent collaborative mechanism, the problems of slow response speed and low accuracy are solved, and a more natural and accurate response effect is achieved.

CN123123596APending Publication Date: 2026-11-17JD DIGITS HAIYI INFORMATION TECHNOLOGY CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510639354.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-05-16
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

In live streaming rooms, existing single intelligent agents or response methods based on preset rules are slow to respond and have low accuracy when facing complex and real-time changing live streaming environments, resulting in poor response effects.

Method used

A multi-agent collaborative mechanism is adopted. By acquiring the target user's questions and various live broadcast information, the system uses the live broadcast information to acquire an intelligent agent set, a multi-turn dialogue intelligent agent, and a question-response intelligent agent to generate accurate response information.

Benefits of technology

It improves response speed and accuracy, generates more natural and precise response information, and adapts to the complexity and real-time changes of the live broadcast room.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123123596A_ABST
    Figure CN123123596A_ABST
Patent Text Reader

Abstract

This disclosure presents embodiments of a question-and-answer method, apparatus, device, and medium applied to a live-streaming room. One specific implementation of the method includes: acquiring a target question raised by a target user in a target live-streaming room; acquiring a set of intelligent agents using live-streaming information to obtain multi-faceted live-streaming information corresponding to the target live-streaming room; inputting the target question and the multi-faceted live-streaming information into a multi-turn dialogue intelligent agent to generate user intent recognition information corresponding to the target user; inputting the user intent recognition information and the multi-faceted live-streaming information into a question-and-answer intelligent agent to generate response information; and replying to the target question with the response information within the target live-streaming room. This implementation relates to the use of intelligent agents, utilizing a multi-agent collaborative mechanism to integrate the advantages of multiple agents, thereby generating more natural response information efficiently and accurately.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The embodiments disclosed herein relate to the field of computer technology, and more specifically to a question-and-answer method, apparatus, device, and medium applied in a live streaming room. Background Technology

[0002] Currently, with the continuous development of large-scale model technology, its application in the live streaming field is becoming increasingly widespread. The typical approach to responding to user questions in a live stream is either based on a single intelligent agent or on a pre-defined rule-based response method.

[0003] However, the inventors discovered that the following technical problems often arise when using the above method:

[0004] When faced with the complex and ever-changing live streaming environment, there are often problems such as slow response speed and low accuracy, resulting in poor response results.

[0005] The information disclosed in this background section is only intended to enhance the understanding of the background of the inventive concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0007] Some embodiments of this disclosure provide methods, apparatuses, electronic devices, computer-readable media, and program products for answering questions in live streaming rooms to solve the technical problems mentioned in the background section above.

[0008] In a first aspect, some embodiments of this disclosure provide a question-and-answer method applied to a live-streaming room, comprising: acquiring a target question raised by a target user in a target live-streaming room; acquiring a set of intelligent agents using live-streaming information to acquire multi-faceted live-streaming information corresponding to the target live-streaming room; inputting the target question and the multi-faceted live-streaming information into a multi-turn dialogue agent to generate user intent recognition information corresponding to the target user; inputting the user intent recognition information and the multi-faceted live-streaming information into a question-and-answer agent to generate response information; and replying to the target question in the target live-streaming room with the response information.

[0009] Optionally, the aforementioned set of intelligent agents for acquiring live streaming information includes: a user interaction behavior evaluation agent; the aforementioned multi-faceted live streaming information includes: response optimization strategy information; and the aforementioned use of the set of intelligent agents for acquiring live streaming information to acquire multi-faceted live streaming information corresponding to the target live streaming room includes: acquiring a set of user interaction behavior information corresponding to the target live streaming room; filtering user interaction behavior information corresponding to the target user from the set of user interaction behavior information as target user interaction behavior information; determining the target user's sentiment tendency information and interaction behavior tendency information within the target time period using the aforementioned user interaction behavior information and the aforementioned user interaction behavior evaluation agent; and generating the aforementioned response optimization strategy information based on the aforementioned sentiment tendency information and the aforementioned interaction behavior tendency information.

[0010] Optionally, the aforementioned set of intelligent agents for obtaining live streaming information includes: a live streaming room information obtaining intelligent agent, and the aforementioned multi-faceted live streaming information includes: live streaming room introduction information; and the aforementioned use of the set of intelligent agents for obtaining live streaming information to obtain multi-faceted live streaming information corresponding to the aforementioned target live streaming room includes: using the aforementioned live streaming room information obtaining intelligent agent to obtain the live streaming room introduction information corresponding to the aforementioned target live streaming room.

[0011] Optionally, the aforementioned set of intelligent agents for acquiring live streaming information includes: a knowledge retrieval agent; the aforementioned multi-faceted live streaming information includes: live streaming domain knowledge information; and the aforementioned use of the set of intelligent agents for acquiring live streaming information to acquire multi-faceted live streaming information corresponding to the aforementioned target live streaming room includes: determining the live streaming content domain information corresponding to the aforementioned target live streaming room; and using the aforementioned knowledge retrieval agent to acquire the live streaming domain knowledge information corresponding to the aforementioned live streaming content domain information.

[0012] Optionally, the above-mentioned input of the target question and the above-mentioned multi-faceted live streaming information into the multi-turn dialogue agent to generate user intent recognition information corresponding to the target user includes: obtaining historical dialogue information between the target user and the digital human corresponding to the target live streaming room; inputting at least one of the above-mentioned response optimization strategy information, the above-mentioned live streaming room introduction information and the above-mentioned live streaming domain knowledge information, the above-mentioned target question and the above-mentioned historical dialogue information into the multi-turn dialogue agent to generate the above-mentioned user intent recognition information.

[0013] Optionally, the above-mentioned inputting at least one of the above-mentioned response optimization strategy information, the above-mentioned live room introduction information, and the above-mentioned live room domain knowledge information, the above-mentioned target question, and the above-mentioned historical dialogue information into the above-mentioned multi-turn dialogue agent to generate the above-mentioned user intent recognition information includes: obtaining response requirement information corresponding to the above-mentioned response optimization strategy information; and inputting the above-mentioned response optimization strategy information, the above-mentioned live room introduction information, the above-mentioned live room domain knowledge information, the above-mentioned response requirement information, the above-mentioned target question, and the above-mentioned historical dialogue information into the above-mentioned multi-turn dialogue agent to generate the above-mentioned user intent recognition information.

[0014] Optionally, the above-mentioned inputting the response optimization strategy information, the live broadcast room introduction information, the live broadcast domain knowledge information, the response requirement information, the target question, and the historical dialogue information into the multi-turn dialogue agent to generate the user intent recognition information includes: using the multi-turn dialogue agent to perform a question semantic summary on the target question to generate question semantic summary information; and inputting the above-mentioned response optimization strategy information, the live broadcast room introduction information, the live broadcast domain knowledge information, the response requirement information, the target question, the question semantic summary information, and the historical dialogue information into the multi-turn dialogue agent to generate the user intent recognition information.

[0015] Secondly, some embodiments of this disclosure provide a question-and-answer device for use in a live streaming room, comprising: a first acquisition unit configured to acquire a target question raised by a target user in a target live streaming room; a second acquisition unit configured to acquire a set of intelligent agents using live streaming information to acquire multi-faceted live streaming information corresponding to the target live streaming room; a first generation unit configured to input the target question and the multi-faceted live streaming information into a multi-turn dialogue intelligent agent to generate user intent recognition information corresponding to the target user; a second generation unit configured to input the user intent recognition information and the multi-faceted live streaming information into a question-and-answer intelligent agent to generate response information; and a reply unit configured to reply to the target question in the target live streaming room using the response information.

[0016] Optionally, the aforementioned set of intelligent agents for acquiring live streaming information includes: a user interaction behavior evaluation intelligent agent; the aforementioned multi-faceted live streaming information includes: response optimization strategy information; and the second acquisition unit can be configured to: acquire the user interaction behavior information set corresponding to the aforementioned target live streaming room; filter out the user interaction behavior information corresponding to the aforementioned target user from the aforementioned user interaction behavior information set, as the target user interaction behavior information; based on the aforementioned target user interaction behavior information, use the aforementioned user interaction behavior evaluation intelligent agent to determine the sentiment tendency information and interaction behavior tendency information of the aforementioned target user within the target time period; and generate the aforementioned response optimization strategy information based on the aforementioned sentiment tendency information and the aforementioned interaction behavior tendency information.

[0017] Optionally, the aforementioned set of intelligent agents for obtaining live streaming information includes: a live streaming room information obtaining intelligent agent, and the aforementioned multi-faceted live streaming information includes: live streaming room introduction information; and the second obtaining unit can be configured to: use the aforementioned live streaming room information obtaining intelligent agent to obtain the live streaming room introduction information corresponding to the aforementioned target live streaming room.

[0018] Optionally, the aforementioned set of intelligent agents for acquiring live streaming information includes: a knowledge retrieval agent; the aforementioned multi-faceted live streaming information includes: live streaming domain knowledge information; and the second acquisition unit can be configured to: determine the live streaming content domain information corresponding to the aforementioned target live streaming room; and use the aforementioned knowledge retrieval agent to acquire the live streaming domain knowledge information corresponding to the aforementioned live streaming content domain information.

[0019] Optionally, the first generation unit can be configured to: acquire historical dialogue information between the target user and the digital human corresponding to the target live room; input at least one of the response optimization strategy information, the live room introduction information, and the live domain knowledge information, the target question, and the historical dialogue information into the multi-turn dialogue agent to generate the user intent recognition information.

[0020] Optionally, the first generation unit can be configured to: obtain the response requirement information corresponding to the above response optimization strategy information; input the above response optimization strategy information, the above live room introduction information, the above live domain knowledge information, the above response requirement information, the above target question, and the above historical dialogue information into the above multi-turn dialogue agent to generate the above user intent recognition information.

[0021] Optionally, the first generation unit can be configured to: utilize the aforementioned multi-turn dialogue agent to perform a semantic summary of the target question to generate question semantic summary information; input the aforementioned response optimization strategy information, the aforementioned live room introduction information, the aforementioned live domain knowledge information, the aforementioned response requirement information, the aforementioned target question, the aforementioned question semantic summary information, and the aforementioned historical dialogue information into the aforementioned multi-turn dialogue agent to generate the aforementioned user intent recognition information.

[0022] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, such that when the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any implementation of the first aspect.

[0023] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method as described in any implementation of the first aspect.

[0024] Fifthly, some embodiments of this disclosure provide a computer program product, including a computer program that, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0025] The above embodiments of this disclosure have the following beneficial effects: By using the question-and-response model training method for live streaming rooms according to some embodiments of this disclosure, and leveraging the multi-agent collaborative mechanism to integrate the advantages of multiple agents, more natural response information can be generated efficiently and accurately. Specifically, the reason why the related response information is not accurate and efficient is that, when facing the complex and real-time changing live streaming environment, there are often problems with slow response speed and low accuracy, resulting in poor response effects. Based on this, the question-and-response method for live streaming rooms according to some embodiments of this disclosure first obtains the target question raised by the target user in the target live streaming room, so as to generate response information for the target question. Then, by using the live streaming information acquisition agent set, the multi-faceted live streaming information corresponding to the target live streaming room can be obtained accurately and efficiently. Here, by using the live streaming information acquisition agent set, collaborative processing between multiple agents can be realized, and the advantages of each agent can be integrated to generate multi-faceted live streaming information accurately and effectively. In addition, by obtaining multi-faceted live streaming information, it is convenient to consider more comprehensive live streaming information in the subsequent process of generating user intent recognition information and response information, ensuring the accuracy of the generation of user intent recognition information and response information. Next, the aforementioned user intent identification information and the aforementioned multi-faceted live broadcast information are input into the question-response agent to accurately generate response information. Finally, the target question is answered with the aforementioned response information within the target live broadcast room. In summary, by coordinating the work among the live broadcast information acquisition agent set, the multi-turn dialogue agent, and the question-response agent, the response speed is improved, enabling accurate and efficient responses to the target question and the generation of response information. Furthermore, by considering multi-faceted live broadcast information, user intent identification information and response information can be accurately generated. Attached Figure Description

[0026] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0027] Figure 1 This is a schematic diagram illustrating an application scenario of a question-and-answer method for a live streaming room, based on some embodiments of this disclosure.

[0028] Figure 2 This is a flowchart of some embodiments of a question-and-answer method applied to a live streaming room according to the present disclosure;

[0029] Figure 3 This is a flowchart of some other embodiments of the question-and-answer method applied to a live broadcast room according to the present disclosure;

[0030] Figure 4 This is a schematic diagram of the structure of some embodiments of a question-and-answer device applied in a live broadcast room according to the present disclosure;

[0031] Figure 5 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure. Detailed Implementation

[0032] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0033] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0034] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0035] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0036] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0037] Before performing any of the operations involving the collection, storage, or use of user personal information (such as response information) disclosed in this disclosure, the relevant organizations or individuals shall fulfill their obligations, including conducting personal information security impact assessments, informing personal information subjects, and obtaining prior authorization and consent from personal information subjects.

[0038] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0039] Figure 1 This is a schematic diagram of an application scenario of a question-and-answer method applied to a live broadcast room according to some embodiments of the present disclosure.

[0040] exist Figure 1In this application scenario, firstly, electronic device 101 can acquire the target question 103 raised by the target user in the target live streaming room 102. In this application scenario, the target live streaming room 102 can be a live streaming room for selling products to digital humans. The target question 103 can be "What is the value of product A?". Then, electronic device 101 can use the live streaming information acquisition intelligent agent set 104 to acquire multi-faceted live streaming information 105 corresponding to the target live streaming room 102. In this application scenario, the live streaming information acquisition intelligent agent set 104 can include: live streaming information acquisition intelligent agent 1, live streaming information acquisition intelligent agent 2, and live streaming information acquisition intelligent agent 3. The multi-faceted live streaming information 105 includes: information related to product A, information related to product B, information related to product C, the live streaming room title, and information about the digital human in the live streaming room. Next, electronic device 101 can input the target question 103 and the multi-faceted live streaming information 105 into the multi-turn dialogue intelligent agent 106 to generate user intent recognition information 107 corresponding to the target user. In this application scenario, the user intent recognition information 107 can be "Buy product A". Furthermore, the electronic device 101 can input the aforementioned user intent recognition information 107 and the aforementioned multi-faceted live broadcast information 105 into the question-response intelligent agent 108 to generate response information 109. In this application scenario, response information 109 could be "Product A is priced at 15 yuan, a favorable price, and of good quality." Finally, the electronic device 101 can reply to the aforementioned target question 103 within the aforementioned target live broadcast room 102 using the aforementioned response information 109.

[0041] It should be noted that the aforementioned electronic device 101 can be either hardware or software. When the electronic device is hardware, it can be implemented as a distributed cluster consisting of multiple servers or terminal devices, or as a single server or a single terminal device. When the electronic device is software, it can be installed in the hardware devices listed above. It can be implemented as, for example, multiple software programs or software modules used to provide distributed services, or as a single software program or software module. No specific limitations are made here.

[0042] It should be understood that Figure 1 The number of electronic devices shown is merely illustrative. Any number of electronic devices can be used depending on the implementation requirements.

[0043] Continue to refer to Figure 2 The flowchart 200 illustrates some embodiments of a question-and-answer method applied to a live streaming room according to the present disclosure. This question-and-answer method applied to a live streaming room includes the following steps:

[0044] Step 201: Obtain the target questions raised by the target users in the target live stream room.

[0045] In some embodiments, the execution entity of the above-described question-and-answer method applied to the live broadcast room (e.g.) Figure 1 The electronic device 101 shown can acquire target questions raised by target users in a target live stream room via wired or wireless connection. The target user can be a user within the target live stream room who raises the target question. The target live stream room can be a live stream room hosting various live stream activities. For example, a live stream activity could be live-streaming e-commerce or a talent show. The target question can be a question entered by the target user in the target live stream room, to be answered by a digital human. The live stream object corresponding to the target live stream room is an object in the form of a digital human.

[0046] Step 202: Use live streaming information to acquire intelligent entities and obtain various aspects of live streaming information corresponding to the target live streaming room.

[0047] In some embodiments, the aforementioned executing entity can utilize a set of live streaming information acquisition agents to acquire multi-faceted live streaming information corresponding to the target live streaming room. The live streaming information acquisition agent can be an agent used to acquire live streaming-related information. There is a one-to-one correspondence between the live streaming information acquisition agents in the set and the live streaming content in the multi-faceted live streaming information. The multi-faceted live streaming information can include live streaming content related to multiple aspects of the target live streaming room. That is, different live streaming information acquisition agents can be used to acquire live streaming content across multiple aspects. For example, multi-faceted live streaming information includes: live streaming smoothness, live streaming audience information, live streaming topic, and live streaming discussion information. The corresponding set of live streaming information acquisition agents can include: a live streaming information acquisition agent for acquiring live streaming smoothness, a live streaming information acquisition agent for acquiring live streaming audience information, a live streaming information acquisition agent for acquiring the live streaming topic, and a live streaming information acquisition agent for acquiring live streaming discussion information.

[0048] In some optional implementations of certain embodiments, the aforementioned set of intelligent agents for acquiring live streaming information includes: a user interaction behavior evaluation agent. The aforementioned multi-faceted live streaming information includes: response optimization strategy information.

[0049] Optionally, the aforementioned executing entity can utilize the live streaming information to acquire intelligent entities and obtain various aspects of live streaming information corresponding to the target live streaming room, which may include the following steps:

[0050] The first step is to obtain the user interaction behavior information set corresponding to the target live stream room. There is a one-to-one correspondence between the user interaction behavior information in this set and the user information in the target live stream room's user information set. The user interaction behavior information can represent the various interactive behaviors of the corresponding user within the target live stream room. This information can be a chain of interactive behaviors with user interactions as nodes. For example, for target user A, the corresponding user interaction behavior information could be: "Entering the live stream room" -> "Multiple rounds of bullet screen interaction" -> "Entering the product details page" -> "Clicking to purchase".

[0051] For example, user interaction behavior information could be:

[0052] 1. Enter the live stream room

[0053] 2. Bullet screen interaction: [

[0054] {role:"user",comment:"*Is there a discount for *50?"},

[0055] {role:"bot",comment:"Are you referring to the r*50 product?"}

[0056] {role:"user",comment:"Yes"},

[0057] {role:"bot",comment:"The prices in our live stream are already discounted; it's rare to find prices like this normally. You can check the shopping bag in the bottom right corner; there's a coupon for r*50. After claiming it, it will be automatically deducted at checkout on the order page."} ]

[0059] 3. See the product details page.

[0060] The second step is to filter out the user interaction behavior information corresponding to the target user from the above user interaction behavior information set, and use it as the target user interaction behavior information.

[0061] The third step involves using the aforementioned target user interaction behavior information, along with the user interaction behavior evaluation agent, to determine the target user's sentiment and interaction behavior tendencies within the target time period. The target time period can be a historical time period within a predetermined duration prior to the current time. Sentiment information characterizes the target user's emotional inclination within the target livestream room. For example, for a livestream room promoting agricultural products, the corresponding sentiment information could be the target user's positive sentiment towards the target agricultural products. Similarly, for a debate livestream room, the corresponding sentiment information could be the target user's agreement with the target viewpoint. Interaction behavior tendencies can be the interactive behaviors the target user might perform within the target livestream room. For example, for a livestream room promoting agricultural products, the interaction behavior tendencies could be the tendency to purchase the target agricultural products. Similarly, for a debate livestream room, the interaction behavior tendencies could be the tendency to like a digital persona holding the target viewpoint.

[0062] As an example, firstly, the aforementioned executing agent can generate prompts representing sentiment tendency information and interaction behavior tendency information within a target time period based on the target user's interaction behavior information. Then, the prompts are input into the aforementioned user interaction behavior evaluation agent to obtain the sentiment tendency information and interaction behavior tendency information.

[0063] The fourth step involves generating the aforementioned response optimization strategy information based on the aforementioned sentiment information and interaction behavior tendency information. This response optimization strategy information optimizes the response to the target question. For example, sentiment information might indicate a positive sentiment towards the target agricultural product, and interaction behavior tendency information might indicate a purchase intention of the target agricultural product. Correspondingly, the response optimization strategy information could be an optimization strategy that adds product introductions and advantages of the target agricultural product to the response information when the target question is determined to be related to the target agricultural product. For example, the response optimization strategy information could be, but is not limited to, one of the following: supplementing information on popular features / parameters, adding user reviews / sales endorsements, proactively asking if additional information is needed, providing more discount combinations, or recommending related accessories.

[0064] As an example, firstly, the aforementioned executing agent can generate strategy generation prompts that represent the generation of response optimization strategy information based on sentiment tendency information, interaction behavior tendency information, and target user interaction behavior information. Then, the strategy generation prompts are input into the user interaction behavior evaluation agent to generate the aforementioned response optimization strategy information.

[0065] In some optional implementations of certain embodiments, the aforementioned set of intelligent agents for obtaining live streaming information includes: a live streaming room information obtaining intelligent agent, and the aforementioned multi-faceted live streaming information includes: live streaming room introduction information. The live streaming room information obtaining intelligent agent may be an intelligent agent that obtains live streaming room introduction information. The live streaming room introduction information may be the introduction information corresponding to the target live streaming room. In practice, the live streaming room introduction information may include, but is not limited to, at least one of the following: live streaming store, live streaming title, live streaming products, live streaming activities, and live streaming likes. The live streaming room information obtaining intelligent agent may support the function of extracting live streaming-related information.

[0066] Optionally, the aforementioned executing entity may use the aforementioned live stream information to acquire intelligent agents and obtain the live stream introduction information corresponding to the aforementioned target live stream.

[0067] In some optional implementations of certain embodiments, the aforementioned set of intelligent agents for acquiring live streaming information includes a knowledge retrieval agent. The aforementioned multi-faceted live streaming information includes live streaming domain knowledge information. The knowledge retrieval agent can be an agent used for knowledge retrieval. That is, the knowledge retrieval agent includes a large language model for retrieving professional knowledge corresponding to the target domain. The live streaming domain knowledge information can be domain-specific professional knowledge corresponding to the live streaming domain.

[0068] Optionally, the aforementioned executing entity may utilize the live streaming information acquisition intelligent body set to obtain various aspects of live streaming information corresponding to the target live streaming room, including the following steps:

[0069] The first step is to determine the content domain information corresponding to the target livestream room. This content domain information can be the domain information related to the livestream content of the target livestream room. For example, for a target livestream room selling agricultural products, the corresponding content domain information could be related to the agricultural products domain.

[0070] The second step is to use the aforementioned knowledge retrieval agent to obtain the live streaming domain knowledge information corresponding to the aforementioned live streaming content domain information.

[0071] As an example, if the information in the live broadcast content is related to agricultural products, the aforementioned implementing entity can use a knowledge retrieval agent to obtain the professional knowledge of agricultural products corresponding to the information in the agricultural product field, and use it as knowledge information in the live broadcast field.

[0072] As an example, firstly, the generation representation generates domain-specific knowledge retrieval prompts based on the domain information of the live stream content. Then, the knowledge retrieval prompts are input into the large language model corresponding to the knowledge retrieval agent to generate live stream domain knowledge information.

[0073] Step 203: Input the aforementioned target question and the aforementioned multi-faceted live broadcast information into the multi-turn dialogue agent to generate user intent recognition information corresponding to the aforementioned target user.

[0074] In some embodiments, the aforementioned executing entity can input the target question and the aforementioned multi-faceted live broadcast information into a multi-turn dialogue agent to generate user intent recognition information corresponding to the target user. The multi-turn dialogue agent can be an agent that performs multi-turn dialogue operations. That is, the multi-turn dialogue agent can engage in multi-turn dialogues with the live broadcast audience (i.e., the user) to identify the audience's intent. The user intent recognition information can be the identified intent of the user in asking the target question. In practice, the multi-turn dialogue agent can be a large language model deployed to support information extraction, question-and-answer operations, and intent recognition. Information extraction can involve extracting key live broadcast information from various aspects of live broadcast information for subsequent auxiliary understanding of the user intent recognition information corresponding to the target user. Question-and-answer operations can be used to conduct question-and-answer operations with the live broadcast audience to further obtain more comprehensive intent information from the audience. Intent recognition can identify the user's intent.

[0075] In some optional implementations of certain embodiments, the execution entity inputs the target question and the multi-faceted live broadcast information into a multi-turn dialogue agent to generate user intent recognition information corresponding to the target user, including the following steps:

[0076] The first step is to obtain the historical dialogue information between the target user and the corresponding digital human in the target live stream. This historical dialogue information can be the content of the conversations between the target user and the digital human within the live stream.

[0077] The second step involves inputting at least one of the above-mentioned response optimization strategy information, the above-mentioned live room introduction information, and the above-mentioned live room domain knowledge information, the above-mentioned target question, and the above-mentioned historical dialogue information into the above-mentioned multi-turn dialogue agent to generate the above-mentioned user intent recognition information.

[0078] As an example, the aforementioned execution entity can input the aforementioned response optimization strategy information, the aforementioned live room introduction information, the aforementioned live room domain knowledge information, the aforementioned target question, and the aforementioned historical dialogue information into the aforementioned multi-turn dialogue agent to generate the aforementioned user intent recognition information.

[0079] As another example, firstly, the aforementioned executing agent can generate a first intent recognition prompt that represents user intent recognition information generated based on at least one of the aforementioned response optimization strategy information, the aforementioned live room introduction information, and the aforementioned live room domain knowledge information, the aforementioned target question, and the aforementioned historical dialogue information. Then, the aforementioned first intent recognition prompt is input into the large language model included in the multi-turn dialogue agent to generate user intent recognition information.

[0080] In some optional implementations of certain embodiments, the execution entity may input at least one of the above-mentioned response optimization strategy information, the above-mentioned live room introduction information, and the above-mentioned live room domain knowledge information, the above-mentioned target question, and the above-mentioned historical dialogue information into the above-mentioned multi-turn dialogue agent to generate the above-mentioned user intent recognition information, including the following steps:

[0081] The first step is to obtain the response requirement information. This information can be various aspects of the response process, including tone of voice, timbre, nickname, intonation, and word choice. Corresponding response requirement information can include: tone of voice requirements, timbre requirements, nickname requirements, intonation requirements, and word choice requirements.

[0082] The second step involves inputting the aforementioned response optimization strategy information, the aforementioned live stream introduction information, the aforementioned live stream domain knowledge information, the aforementioned response requirement information, the aforementioned target question, and the aforementioned historical dialogue information into the aforementioned multi-turn dialogue agent to generate the aforementioned user intent recognition information.

[0083] As an example, firstly, the aforementioned executing entity can generate a second intent recognition prompt that represents user intent recognition information generated based on the aforementioned response optimization strategy information, the aforementioned live broadcast room introduction information, the aforementioned live broadcast domain knowledge information, the aforementioned response requirement information, the aforementioned target question, and the aforementioned historical dialogue information. Then, the aforementioned second intent recognition prompt is input into the large language model included in the multi-turn dialogue agent to generate user intent recognition information.

[0084] Here, by adding response requirement information, multi-turn dialogue agents can output user intent recognition information and response information that meet the response requirements.

[0085] Step 204: Input the above-mentioned user intent identification information and the above-mentioned multi-faceted live broadcast information into the question response agent to generate response information.

[0086] In some embodiments, the aforementioned execution entity may input the aforementioned user intent identification information and the aforementioned multi-faceted live streaming information into the question-response intelligent agent to generate response information. The question-response intelligent agent may be an agent that generates response information corresponding to a question. The question-response intelligent agent may generate corresponding response information based on the user intent identification information. In practice, the response information may be multimodal information. For example, the response information may be response text or response audio. That is, the response information is information that answers the target question.

[0087] Step 205: Reply to the target question in the target live broadcast room using the above response information.

[0088] In some embodiments, the aforementioned executing entity may respond to the aforementioned target question with the aforementioned response information within the aforementioned target live broadcast room.

[0089] As an example, the aforementioned entity may respond to the aforementioned target question in the aforementioned target live broadcast room using the aforementioned response information in the form of a text reply or an audio reply.

[0090] The above embodiments of this disclosure have the following beneficial effects: By using the question-and-response model training method for live streaming rooms according to some embodiments of this disclosure, and leveraging the multi-agent collaborative mechanism to integrate the advantages of multiple agents, more natural response information can be generated efficiently and accurately. Specifically, the reason why the related response information is not accurate and efficient is that, when facing the complex and real-time changing live streaming environment, there are often problems with slow response speed and low accuracy, resulting in poor response effects. Based on this, the question-and-response method for live streaming rooms according to some embodiments of this disclosure first obtains the target question raised by the target user in the target live streaming room, so as to generate response information for the target question. Then, by using the live streaming information acquisition agent set, the multi-faceted live streaming information corresponding to the target live streaming room can be obtained accurately and efficiently. Here, by using the live streaming information acquisition agent set, collaborative processing between multiple agents can be realized, and the advantages of each agent can be integrated to generate multi-faceted live streaming information accurately and effectively. In addition, by obtaining multi-faceted live streaming information, it is convenient to consider more comprehensive live streaming information in the subsequent process of generating user intent recognition information and response information, ensuring the accuracy of the generation of user intent recognition information and response information. Next, the aforementioned user intent identification information and the aforementioned multi-faceted live broadcast information are input into the question-response agent to accurately generate response information. Finally, the target question is answered with the aforementioned response information within the target live broadcast room. In summary, by coordinating the work among the live broadcast information acquisition agent set, the multi-turn dialogue agent, and the question-response agent, the response speed is improved, enabling accurate and efficient responses to the target question and the generation of response information. Furthermore, by considering multi-faceted live broadcast information, user intent identification information and response information can be accurately generated.

[0091] Further reference Figure 3 The diagram illustrates a flow 300 of another embodiment of a question-and-answer method applied to a live streaming room according to the present disclosure. This question-and-answer method for a live streaming room includes the following steps:

[0092] Step 301: Obtain the target questions raised by the target users in the target live stream room.

[0093] Step 302: Use live streaming information to acquire intelligent entities and obtain various aspects of live streaming information corresponding to the target live streaming room.

[0094] Step 303: Using the aforementioned multi-turn dialogue agent, perform a semantic summary of the target question to generate semantic summary information.

[0095] In some embodiments, the executing entity (e.g. Figure 1 The electronic device 101 shown can utilize the aforementioned multi-turn dialogue agent to perform a semantic summary of the target question, thereby generating semantic summary information. This semantic summary information can be information summarizing the semantic content corresponding to the target question. That is, the semantic summary information includes the key semantic content corresponding to the target question.

[0096] Step 304: Input the above-mentioned response optimization strategy information, the above-mentioned live room introduction information, the above-mentioned live room domain knowledge information, the above-mentioned response requirement information, the above-mentioned target question, the above-mentioned question semantic summary information, and the above-mentioned historical dialogue information into the above-mentioned multi-turn dialogue agent to generate the above-mentioned user intent recognition information.

[0097] In some embodiments, the execution entity may input the response optimization strategy information, the live room introduction information, the live domain knowledge information, the response requirement information, the target question, the question semantic summary information, and the historical dialogue information into the multi-turn dialogue agent to generate the user intent recognition information.

[0098] Step 305: Input the above-mentioned user intent identification information and the above-mentioned multi-faceted live broadcast information into the question response agent to generate response information.

[0099] Step 306: Reply to the target question in the target live broadcast room using the above response information.

[0100] In some embodiments, the specific implementation of steps 301-302 and 305-306 and their resulting technical effects can be found in [reference needed]. Figure 2 Steps 201-202 and 204-205 in the corresponding embodiments will not be repeated here.

[0101] from Figure 3 It can be seen from this that, with Figure 2 Compared to the description of some corresponding embodiments, Figure 3 In some corresponding embodiments, the process 300 of the question response method applied to the live broadcast room adds question semantic summary information so that the multi-turn dialogue agent can generate more accurate user intent recognition information based on the above-mentioned response optimization strategy information, the above-mentioned live broadcast room introduction information, the above-mentioned live broadcast domain knowledge information, the above-mentioned response requirement information, the above-mentioned target question, the above-mentioned question semantic summary information and the above-mentioned historical dialogue information.

[0102] Further reference Figure 4 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a question-and-answer device applied to a live broadcast room. These device embodiments are similar to... Figure 2 Corresponding to the method embodiments shown, this question-and-answer device for live streaming can be specifically applied to various electronic devices.

[0103] like Figure 4 As shown, a question-and-answer device 400 applied in a live streaming room includes: a first acquisition unit 401, a second acquisition unit 402, a first generation unit 403, a second generation unit 404, and a response unit 405. The first acquisition unit 401 is configured to acquire a target question raised by a target user in a target live streaming room; the second acquisition unit 402 is configured to acquire multi-faceted live streaming information corresponding to the target live streaming room using a live streaming information acquisition intelligence set; the first generation unit 403 is configured to input the target question and the multi-faceted live streaming information into a multi-turn dialogue intelligence to generate user intent recognition information corresponding to the target user; the second generation unit 404 is configured to input the user intent recognition information and the multi-faceted live streaming information into a question-and-answer intelligence to generate response information; and the response unit 405 is configured to respond to the target question in the target live streaming room using the response information.

[0104] In some optional implementations of certain embodiments, the aforementioned set of intelligent agents for acquiring live streaming information includes: a user interaction behavior evaluation intelligent agent; the aforementioned multi-faceted live streaming information includes: response optimization strategy information; and the second acquisition unit 402 may be further configured to: acquire the user interaction behavior information set corresponding to the target live streaming room; filter out the user interaction behavior information corresponding to the target user from the aforementioned user interaction behavior information set as the target user interaction behavior information; determine the emotional tendency information and interaction behavior tendency information of the target user within the target time period using the aforementioned user interaction behavior evaluation intelligent agent based on the aforementioned target user interaction behavior information; and generate the aforementioned response optimization strategy information based on the aforementioned emotional tendency information and the aforementioned interaction behavior tendency information.

[0105] In some optional implementations of some embodiments, the above-mentioned set of intelligent agents for obtaining live information includes: a live room information obtaining intelligent agent, and the above-mentioned multi-faceted live information includes: live room introduction information; and the second obtaining unit 402 can be further configured to: use the above-mentioned live room information obtaining intelligent agent to obtain the live room introduction information corresponding to the target live room.

[0106] In some optional implementations of some embodiments, the above-mentioned live information acquisition intelligent agent set includes: a knowledge retrieval intelligent agent, the above-mentioned multi-faceted live information includes: live domain knowledge information; and the second acquisition unit 402 can be further configured to: determine the live content domain information corresponding to the above-mentioned target live room; and use the above-mentioned knowledge retrieval intelligent agent to acquire the live domain knowledge information corresponding to the above-mentioned live content domain information.

[0107] In some optional implementations of some embodiments, the first generation unit 403 may be further configured to: acquire historical dialogue information between the target user and the digital human corresponding to the target live room; input at least one of the response optimization strategy information, the live room introduction information and the live domain knowledge information, the target question and the historical dialogue information into the multi-turn dialogue agent to generate the user intent recognition information.

[0108] In some optional implementations of some embodiments, the first generation unit 403 may be further configured to: utilize the aforementioned multi-turn dialogue agent to perform a question semantic summary on the aforementioned target question to generate question semantic summary information; input the aforementioned response optimization strategy information, the aforementioned live room introduction information, the aforementioned live domain knowledge information, the aforementioned response requirement information, the aforementioned target question, the aforementioned question semantic summary information, and the aforementioned historical dialogue information to the aforementioned multi-turn dialogue agent to generate the aforementioned user intent recognition information.

[0109] It is understandable that the units described in the question-and-answer device 400 used in the live broadcast room are similar to those in the reference. Figure 2 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the question-and-answer device 400 applied to the live broadcast room and the units contained therein, and will not be repeated here.

[0110] The following is for reference. Figure 5 It illustrates electronic devices suitable for implementing some embodiments of this disclosure (e.g., Figure 1 A schematic diagram of the structure of electronic device 101)500. Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0111] like Figure 5As shown, the electronic device 500 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory 502 or a program loaded from a storage device 508 into a random access memory 503. The random access memory 503 also stores various programs and data required for the operation of the electronic device 500. The processing unit 501, the read-only memory 502, and the random access memory 503 are interconnected via a bus 504. An input / output interface 505 is also connected to the bus 504.

[0112] Typically, the following devices can be connected to the input / output interface 505: input devices 506 including, for example, a touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 507 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 508 including, for example, magnetic tape, hard disk, etc.; and communication devices 509. Communication device 509 allows electronic device 500 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 An electronic device 500 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 5 Each box shown can represent a device or multiple devices as needed.

[0113] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 509, or installed from a storage device 508, or installed from a read-only memory 502. When the computer program is executed by the processing device 501, it performs the functions defined above in the methods of some embodiments of this disclosure.

[0114] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0115] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0116] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, cause the electronic device to: acquire a target question raised by a target user in a target live stream; acquire a set of intelligent agents using live stream information to obtain multi-faceted live stream information corresponding to the target live stream; input the target question and the multi-faceted live stream information into a multi-turn dialogue intelligent agent to generate user intent recognition information corresponding to the target user; input the user intent recognition information and the multi-faceted live stream information into a question-response intelligent agent to generate response information; and respond to the target question in the target live stream using the response information.

[0117] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed 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 remote computers, the remote computer can 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 can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0118] 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 this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated 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, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0119] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including a first acquisition unit, a second acquisition unit, a first generation unit, a second generation unit, and a response unit. The names of these units do not necessarily limit the specific unit; for example, the first acquisition unit may also be described as "a unit that acquires the target questions raised by the target user in the target live stream room."

[0120] The functions described above in this document can be performed, at least in part, by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0121] Some embodiments of this disclosure also provide a computer program product, including a computer program that, when executed by a processor, implements any of the above-described question-and-answer methods applied to a live broadcast room.

[0122] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A question-and-answer method for use in live streaming, comprising: Obtain the target questions raised by target users in the target live stream; The intelligent body is acquired by utilizing live streaming information to obtain various aspects of live streaming information corresponding to the target live streaming room. The target question and the multi-faceted live broadcast information are input into a multi-turn dialogue agent to generate user intent recognition information corresponding to the target user; The user intent recognition information and the multi-faceted live broadcast information are input into the question response agent to generate response information; Respond to the target question with the response information within the target live stream room.

2. The method according to claim 1, wherein, The set of intelligent agents for acquiring live streaming information includes: a user interaction behavior evaluation agent; and the multi-faceted live streaming information includes: response optimization strategy information; and The method of acquiring intelligent entities using live streaming information to obtain multi-faceted live streaming information corresponding to the target live streaming room includes: Obtain the set of user interaction behavior information corresponding to the target live streaming room; The user interaction behavior information corresponding to the target user is filtered from the user interaction behavior information set and used as the target user interaction behavior information. Based on the target user's interaction behavior information, the intelligent agent is evaluated using the user interaction behavior to determine the target user's emotional tendency information and interaction behavior tendency information within the target time period. The response optimization strategy information is generated based on the emotional tendency information and the interaction behavior tendency information.

3. The method according to claim 1, wherein, The set of intelligent agents for acquiring live streaming information includes: an intelligent agent for acquiring live streaming room information, and the various aspects of live streaming information include: live streaming room introduction information; and The method of acquiring intelligent entities using live streaming information to obtain multi-faceted live streaming information corresponding to the target live streaming room includes: The intelligent agent is obtained using the live stream information to acquire the live stream introduction information corresponding to the target live stream.

4. The method according to claim 1, wherein, The set of intelligent agents for acquiring live streaming information includes: a knowledge retrieval intelligent agent; and the multi-faceted live streaming information includes: live streaming domain knowledge information; and The method of acquiring intelligent entities using live streaming information to obtain multi-faceted live streaming information corresponding to the target live streaming room includes: Determine the live content domain information corresponding to the target live room; The knowledge retrieval agent is used to obtain live streaming domain knowledge information corresponding to the live streaming content domain information.

5. The method according to any one of claims 1-4, wherein, The step of inputting the target question and the multi-faceted live broadcast information into a multi-turn dialogue agent to generate user intent recognition information corresponding to the target user includes: Obtain historical dialogue information between the target user and the digital person corresponding to the target live stream room; The response optimization strategy information, the live room introduction information, and the live room domain knowledge information, the target question, and the historical dialogue information are input into the multi-turn dialogue agent to generate the user intent recognition information.

6. The method according to claim 5, wherein, The step of inputting at least one of the response optimization strategy information, the live room introduction information, and the live room domain knowledge information, the target question, and the historical dialogue information into the multi-turn dialogue agent to generate the user intent recognition information includes: Obtain the response requirement information corresponding to the response optimization strategy information; The response optimization strategy information, the live room introduction information, the live room domain knowledge information, the response requirement information, the target question, and the historical dialogue information are input into the multi-turn dialogue agent to generate the user intent recognition information.

7. The method according to claim 6, wherein, The step of inputting the response optimization strategy information, the live stream introduction information, the live stream domain knowledge information, the response requirement information, the target question, and the historical dialogue information into the multi-turn dialogue agent to generate the user intent recognition information includes: Using the multi-turn dialogue agent, the target question is semantically summarized to generate semantic summary information. The response optimization strategy information, the live room introduction information, the live room domain knowledge information, the response requirement information, the target question, the question semantic summary information, and the historical dialogue information are input into the multi-turn dialogue agent to generate the user intent recognition information.

8. A question-and-answer device for use in a live streaming room, comprising: The first acquisition unit is configured to acquire the target question raised by the target user in the target live broadcast room; The second acquisition unit is configured to acquire intelligent entities using live streaming information to acquire various aspects of live streaming information corresponding to the target live streaming room. The first generation unit is configured to input the target question and the multi-faceted live broadcast information into a multi-turn dialogue agent to generate user intent recognition information corresponding to the target user. The second generation unit is configured to input the user intent recognition information and the multi-faceted live broadcast information into the question response agent to generate response information; The response unit is configured to respond to the target question with the response information within the target live stream room.

9. An electronic device, comprising: One or more processors; Storage device, on which one or more programs are stored, When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1-7.

10. A computer-readable medium having a computer program stored thereon, wherein, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-7.

11. A computer program product comprising a computer program that, when executed by a processor, implements the method according to any one of claims 1-7.