Intelligent agent message sending method and device based on asynchronous communication group, equipment and medium

Abstract

The invention discloses an agent message sending method and device based on an asynchronous communication group, equipment and a medium, and relates to the technical field of natural language processing, and the method comprises the steps: determining the speaking frequency of an agent in a current time period of the asynchronous communication group; the asynchronous communication group is a group related to the online team conference; generating current speaking scheduling information according to a matching relationship between the speaking frequency and a preset frequency threshold value; the preset frequency threshold is a threshold constructed based on the number of real-time online members of the asynchronous communication group; and based on the current speaking scheduling information, determining whether to combine the messages sent by the group members in the current time period, and generating a to-be-sent message, so that the to-be-sent message is sent to the asynchronous communication group through the intelligent agent when the to-be-sent message is generated. Therefore, the speaking opportunity of the agents can be determined according to the speaking frequency of the agents in the asynchronous communication group, the speaking opportunity is determined autonomously, and the method can adapt to asynchronous communication scenes such as online team conferences and the like.

Description

Technical Field

[0001] The present invention relates to the field of natural language processing technology, and in particular to a method, device, equipment and medium for sending intelligent messages based on an asynchronous communication group. Background Art

[0002] Most currently used applications based on LLMs (Large Language Models) focus on synchronous communication scenarios, such as turn-based conversations between humans and machines. However, many real-world communication scenarios, such as online team meetings, social games, and group chats, are inherently asynchronous. In such asynchronous communication, participants must decide not only what to say but also when to say it. However, research on LLMs in asynchronous group communication environments is relatively scarce. Commonly used social interaction models often presuppose fixed turn times and fail to accurately simulate real-world asynchronous communication scenarios.

[0003] Therefore, how to optimize the speaking behavior of intelligent agents in asynchronous communication groups is a problem to be solved in this field. Summary of the Invention

[0004] In view of this, the present invention aims to provide a method, apparatus, device, and medium for sending messages to agents based on asynchronous communication groups. These methods can determine the speaking timing of agents based on their speaking frequency in the asynchronous communication group, allowing agents to independently decide when to speak, and are adaptable to asynchronous communication scenarios such as online team meetings. The specific solution is as follows:

[0005] In a first aspect, the present application provides a method for sending messages to an agent based on an asynchronous communication group, comprising:

[0006] Determining the speaking frequency of the agent in a current time period of an asynchronous communication group; the asynchronous communication group is a group related to an online team meeting;

[0007] Generating current speech scheduling information corresponding to the agent based on a matching relationship between the speech frequency and a preset frequency threshold; the preset frequency threshold is a threshold constructed based on the number of real-time online members of the asynchronous communication group; group members in the asynchronous communication group include real user members and / or the agent;

[0008] Based on the current speech scheduling information, it is determined whether to combine the messages sent by each group member in the current time period to generate a message to be sent, so that when the message to be sent is generated, the message to be sent is sent to the asynchronous communication group through the agent.

[0009] Optionally, determining the speaking frequency of the agent in the current time period of the asynchronous communication group includes:

[0010] According to the timestamps of the historical messages corresponding to the asynchronous communication group, the messages corresponding to the current time period are updated in real time to obtain a first current message corresponding to the current time period;

[0011] Determining a second current message sent by the agent from the first current message;

[0012] The speaking frequency of the agent in the current time period is determined based on the message quantity relationship between the second current message and the first current message.

[0013] Optionally, generating current speech scheduling information corresponding to the agent according to a matching relationship between the speech frequency and a preset frequency threshold includes:

[0014] If the speaking frequency matches the preset frequency threshold, generating current speaking scheduling information indicating that the agent is allowed to speak;

[0015] If the speaking frequency does not match the preset frequency threshold, current speaking scheduling information indicating that speaking is not allowed is generated for the agent.

[0016] Optionally, if the speaking frequency matches a preset frequency threshold, generating current speaking scheduling information indicating that the agent is allowed to speak includes:

[0017] If the preset frequency threshold is the ratio between the target number and the number of real-time online members, and the speech frequency is less than the preset frequency threshold, it is determined that the speech frequency and the preset frequency threshold are in a matching relationship, and current speech scheduling information representing the allowed speech is generated for the agent; wherein, the target number is 1.

[0018] Optionally, if the speaking frequency matches a preset frequency threshold, generating current speaking scheduling information indicating that the agent is allowed to speak includes:

[0019] If the preset frequency threshold is a threshold for controlling the speaking frequency of the agent, calculated based on the target weight and the number of real-time online members, and the speaking frequency is less than the preset frequency threshold, then it is determined that the speaking frequency and the preset frequency threshold are in a matching relationship, and current speaking scheduling information indicating that the agent is allowed to speak is generated for the agent;

[0020] Among them, the target weight includes a first target weight and a second target weight, the first target weight is a preset weight corresponding to each of the real-time online members, and the second target weight is a real-time weight corresponding to each of the real-time online members; the preset weight is a weight value pre-configured for all group members based on the member identity information of all group members in the asynchronous communication group, and the real-time weight is a weight value calculated in real time for all real-time online members based on the member identity information of all real-time online members in the current asynchronous communication group.

[0021] Optionally, the determining, based on the current speech scheduling information, whether to combine messages sent by each group member within the current time period to generate a message to be sent includes:

[0022] If the current speech scheduling information indicates that speaking is not allowed, then the generation of the message to be sent is prohibited;

[0023] If the current speech scheduling information indicates that speaking is allowed, based on a preset message generation strategy and in combination with messages sent by each of the group members in the current time period, a message to be sent is generated using a preset large language model;

[0024] Among them, the preset message generation strategy is a strategy constructed based on prompt word engineering, word blocking operation and semantic repetition rate to guide the preset large language model to generate messages. The prompt word engineering is a layered prompt word engineering that combines member roles, group communication scenarios and historical messages.

[0025] Optionally, sending the to-be-sent message to the asynchronous communication group by the agent includes:

[0026] Determining the message length of the message to be sent;

[0027] Setting a delay time matching the message length based on a preset manual typing speed;

[0028] The message to be sent is sent to the asynchronous communication group using the delay time through the agent.

[0029] In a second aspect, the present application provides an agent message sending device based on an asynchronous communication group, comprising:

[0030] A speaking frequency determination module is used to determine the speaking frequency of an agent in a current time period of an asynchronous communication group; the asynchronous communication group is a group related to an online team meeting;

[0031] a speech scheduling information generation module, configured to generate current speech scheduling information corresponding to the agent based on a matching relationship between the speech frequency and a preset frequency threshold; the preset frequency threshold being a threshold value constructed based on the number of real-time online members of the asynchronous communication group; group members in the asynchronous communication group including real user members and / or the agent;

[0032] The message generation module is used to determine whether to generate a message to be sent based on the current speech scheduling information in combination with the messages sent by each group member within the current time period, so that when the message to be sent is generated, the message to be sent is sent to the asynchronous communication group through the agent.

[0033] In a third aspect, the present application provides an electronic device, comprising:

[0034] Memory, used to store computer programs;

[0035] A processor is used to execute the computer program to implement the above-mentioned agent message sending method based on asynchronous communication group.

[0036] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the above-mentioned method for sending intelligent messages based on an asynchronous communication group.

[0037] As can be seen from this, the present application can first determine the speaking frequency of the agent in the current time period of the asynchronous communication group; the asynchronous communication group is a group related to online team meetings; then, based on the matching relationship between the speaking frequency and the preset frequency threshold, generate the current speaking scheduling information corresponding to the agent; the preset frequency threshold is a threshold constructed based on the number of real-time online members of the asynchronous communication group; the group members in the asynchronous communication group include real user members and / or the agent; then, based on the current speaking scheduling information, determine whether to generate a to-be-sent message in combination with the messages sent by each group member in the current time period, so that when the to-be-sent message is generated, the to-be-sent message is sent to the asynchronous communication group through the agent. In this way, the present application can determine the speaking timing of the agent based on the speaking frequency of the agent in the asynchronous communication group, can independently determine the speaking timing in an asynchronous communication environment without fixed communication rounds, and can adapt to asynchronous communication scenarios such as online team meetings; and can dynamically adjust the speaking behavior of the agent in combination with the real-time online members of the asynchronous communication group, so that the speaking behavior of the agent can be more adapted to the actual scenario. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0039] Figure 1 This is a flow chart of a method for sending messages to an intelligent agent based on an asynchronous communication group disclosed in this application;

[0040] Figure 2 This is a flowchart of a specific method for sending agent messages based on an asynchronous communication group disclosed in this application;

[0041] Figure 3 A flow chart for judging speech frequency disclosed in this application;

[0042] Figure 4 This is a schematic diagram of the structure of an intelligent message sending device based on an asynchronous communication group disclosed in this application;

[0043] Figure 5 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0044] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0045] See also Figure 1 As shown, the embodiment of the present invention discloses a method for sending messages to an agent based on an asynchronous communication group, comprising:

[0046] Step S11: Determine the speaking frequency of the agent in the current time period of the asynchronous communication group; the asynchronous communication group is a group related to the online team meeting.

[0047] It is understood that the frequency and content of speeches by group members in an asynchronous communication group will influence the agent's message sending decisions. This can consider the agent's speaking frequency within the current time period within the asynchronous communication group; alternatively, it can consider the agent's speaking frequency over the entire time period within the asynchronous communication group, that is, consider the entire history of the asynchronous communication group's messages to determine the agent's speaking frequency. In a specific embodiment, the asynchronous communication group is a group related to online team meetings, but it can also be a group related to scenarios such as social games and online classroom discussions.

[0048] In a specific embodiment, determining the speaking frequency of an agent in a current time period of an asynchronous communication group may include: updating the messages corresponding to the current time period in real time based on the timestamps of historical messages corresponding to the asynchronous communication group to obtain a first current message corresponding to the current time period; determining a second current message sent by the agent from the first current message; and determining the speaking frequency of the agent in the current time period based on the message quantity relationship between the second current message and the first current message. Specifically, for the messages in the current time period of the asynchronous communication group, the messages in the current time period may be updated in real time based on the timestamps of the historical messages and the timestamp of the current time period. The first current message thus obtained is the message sent by each group member (including the agent) in the asynchronous communication group in the current time period. Then, a number of messages sent by the agent may be determined from the first current message to obtain a second current message. Furthermore, based on the ratio between the second current message and the first current message, the speaking frequency of the agent in the current time period may be determined, i.e., the ratio of the number of messages sent by the agent to the total number of messages in the current time period.

[0049] Step S12: Generate the current speech scheduling information corresponding to the intelligent agent based on the matching relationship between the speech frequency and the preset frequency threshold; the preset frequency threshold is a threshold constructed based on the number of real-time online members of the asynchronous communication group; the group members in the asynchronous communication group include real user members and / or the intelligent agent.

[0050] Furthermore, the matching relationship between the agent's speaking frequency and a preset frequency threshold can be considered. Based on this matching relationship, whether the agent should speak can be determined, i.e., current speech scheduling information can be generated. It should be noted that the preset frequency threshold used here can be updated in real time based on the number of real-time online members in the asynchronous communication group. The requirement for the agent's speaking frequency varies with the number of real-time online members. Furthermore, the real-time online members here can include real user members as well as related agents. In specific embodiments, if there are multiple online agents, these multiple online agents can also be classified as real-time online members.

[0051] In a specific embodiment, generating the current speech scheduling information corresponding to the agent based on the matching relationship between the speech frequency and the preset frequency threshold may include: if the speech frequency and the preset frequency threshold are in a matching relationship, generating the current speech scheduling information for the agent indicating that speech is allowed; if the speech frequency and the preset frequency threshold are in a mismatching relationship, generating the current speech scheduling information for the agent indicating that speech is not allowed. Specifically, in the process of generating the current speech scheduling information for the agent based on the matching relationship between the speech frequency of the agent and the preset frequency threshold, the speech frequency and the preset frequency threshold may present a matching relationship or a mismatching relationship, and based on the matching or mismatching relationship, the current speech scheduling information indicating whether speech is allowed or not allowed may be generated for the agent.

[0052] In another specific embodiment, if the speaking frequency matches a preset frequency threshold, generating current speaking schedule information for the agent indicating that it is permitted to speak may include: if the preset frequency threshold is the ratio of the target number to the number of real-time online members, and the speaking frequency is less than the preset frequency threshold, determining that the speaking frequency matches the preset frequency threshold, and generating current speaking schedule information for the agent indicating that it is permitted to speak; wherein the target number is 1. Specifically, the preset frequency threshold can be set in various ways. The frequency threshold can be updated based on the number of real-time online members and the target number of agents (for the current agent, the number is 1), i.e., the ratio between the target number and the number of real-time online members is determined as the preset frequency threshold. Then, determining the relationship between the speaking frequency and the frequency threshold. If the speaking frequency is less than the frequency threshold, it indicates that the agent speaks less, not reaching the average level of online members, and needs to actively participate and speak as much as possible. In this way, determining that the speaking frequency matches the preset frequency threshold, and generating current speaking schedule information for the agent indicating that it is permitted to speak.

[0053] In another specific embodiment, if the speaking frequency matches the preset frequency threshold, generating current speaking scheduling information for the agent indicating that speaking is allowed may include: if the preset frequency threshold is a threshold for controlling the speaking frequency of the agent calculated based on the target weight and the number of real-time online members, and the speaking frequency is less than the preset frequency threshold, then determining that the speaking frequency matches the preset frequency threshold, and generating current speaking scheduling information for the agent indicating that speaking is allowed; wherein the target weight includes a first target weight and a second target weight, the first target weight is the preset weight corresponding to each of the real-time online members, and the second target weight is the real-time weight corresponding to each of the real-time online members; the preset weight is a weight value pre-configured for all group members based on the membership information of all group members in the asynchronous communication group, and the real-time weight is a weight value calculated in real time for all real-time online members based on the membership information of all the real-time online members in the current asynchronous communication group. Specifically, the identity information of group members in an asynchronous communication group varies. Based on this information, corresponding identity weights, i.e., first target weights, can be set. This identity weight also represents the requirement for the speaking frequency of the relevant group members. Furthermore, at different times, the number of real-time online members in the asynchronous communication group can also vary. The first target weight corresponding to each real-time online member is a weight relative to all group members. In this case, the identity information of the real-time online members needs to be considered and the identity weights, i.e., second target weights, recalculated. Accordingly, since the number of online members affects the overall effectiveness of asynchronous communication, a threshold for controlling the speaking frequency of an agent can be calculated in real time, taking into account the first and second target weights of the real-time online members. The relationship between the agent's speaking frequency and the relevant frequency threshold is then determined. If the speaking frequency is less than the relevant frequency threshold, a matching relationship can be determined, and current speaking schedule information indicating that the agent is allowed to speak is generated. Conversely, if the speaking frequency is not less than the relevant frequency threshold, a mismatch can be determined, and current speaking schedule information indicating that speaking is not allowed is generated.

[0054] Step S13: determine whether to generate a message to be sent based on the current speech scheduling information in combination with the messages sent by each group member in the current time period, so that when the message to be sent is generated, the message to be sent is sent to the asynchronous communication group through the agent.

[0055] In this embodiment, the above steps can take into account the speaking frequency of the agent in the current time period, and combine with the preset frequency threshold to generate the current speaking scheduling information corresponding to the agent; it can be understood that the current speaking scheduling information is to allow speaking or not. Furthermore, if the current speaking scheduling information indicates that speaking is not allowed, the generation of the to-be-sent message is prohibited; if the current speaking scheduling information indicates that speaking is allowed, based on the preset message generation strategy, combined with the messages sent by each group member in the current time period, the to-be-sent message can be generated by the preset large language model; the preset message generation strategy involved here is a strategy based on prompt word engineering, word masking operations, and semantic repetition rate, which is used to guide the preset large language model to generate messages, and the prompt word engineering is a layered prompt word engineering that combines member roles, group communication scenarios, and historical messages. It is understandable that group members may go online and offline during the current time period. If a group member A speaks during the current time period, but after sending message a, goes offline before considering whether the intelligent agent speaks, the corresponding real-time online members do not include the group member A who spoke and went offline; however, in the process of the intelligent agent generating the message to be sent, message a needs to be taken into account to ensure the accuracy of the intelligent agent's speech.

[0056] In a specific embodiment, sending the message to be sent to the asynchronous communication group via the agent may include: determining the length of the message to be sent; setting a delay time that matches the message length based on a preset manual typing speed; and sending the message to be sent to the asynchronous communication group via the agent using the delay time. Specifically, in the process of sending the message to be sent to the asynchronous communication group via the agent, a manual typing speed may be simulated, and based on the length of the message to be sent, a delay time that matches the message length may be set for the message to be sent based on the preset manual typing speed. When sending, the message to be sent is sent to the asynchronous communication group using the delay time.

[0057] It can be seen from this that the present application can determine the speaking timing of the intelligent agent based on the speaking frequency of the intelligent agent in the asynchronous communication group, and can independently decide the speaking timing in an asynchronous communication environment without fixed communication rounds. It introduces typing time simulation based on message length, so that the speaking mode of the intelligent agent is closer to the real user, improves the naturalness and realism of communication, and can adapt to asynchronous communication scenarios such as online team meetings; and, it can dynamically adjust the speaking behavior of the intelligent agent in combination with the real-time online members of the asynchronous communication group, so that the speaking behavior of the intelligent agent can be more adapted to the actual scenario.

[0058] like Figure 2 As shown, the embodiment of the present application discloses a method for sending messages to an agent based on an asynchronous communication group, which specifically includes:

[0059] In this embodiment, it is first necessary to clarify the communication environment of the current asynchronous communication group, including chat history, number of active participants, message timestamps, etc.; then, the speaking frequency of the agent is determined based on the ratio of the number of messages sent by the agent in a certain period of time to the total number of messages; the speaking frequency of the agent can also be evaluated based on the entire chat history of the asynchronous communication group.

[0060] Then, based on the agent's speaking frequency, the ratio of the agent's speaking frequency to the average speaking frequency is evaluated, and scheduling prompt words are dynamically generated; Figure 3 In the speaking frequency judgment process shown in the figure, when the speaking frequency f of the agent is less than 1 / n of the number of active participants (i.e., the frequency threshold, where n is the number of active participants), the scheduling prompt encourages the agent to speak more actively; triggering the "encourage speaking" prompt: "Remember to let others hear your voice, so be actively involved and be as active as other participants...", the scheduling module can prompt the LLM to increase the speaking frequency; and when the speaking frequency is higher than the threshold, the prompt is: "Pay attention to the number of messages you send compared to other participants, so that they also have the opportunity to speak...", the scheduling module can prompt the LLM to reduce the speaking frequency.

[0061] Furthermore, the scheduling prompt word can be input into LLM to determine whether to speak ( <send>or <wait>, send or wait), and when the scheduling decision is <send>When sending a message to an asynchronous communication group, the agent needs to input all historical messages from a specific period of time and generate prompt words into the LLM. This allows the agent to retrieve the speech content generated by the LLM, that is, to obtain the message to be sent that matches the current communication scenario. The agent can also perform duplication checks and style matching verification on the message to be sent, such as generating short text (average of 10 words or less), avoiding repetition, matching human language style, prohibiting punctuation, and emphasizing the novelty of the information. It is important to note that when sending a message to an asynchronous communication group, the agent can consider the length of the generated message and calculate the simulated typing time. After a corresponding delay, the message is sent to the group communication environment.

[0062] It is understandable that during the normal operation of the asynchronous communication group, the speaking frequency of the intelligent agent can be counted in real time and compared with the speaking frequency of other participants; based on the comparison results, the content of the scheduling prompt words can be dynamically adjusted to maintain the balance of the overall speaking frequency.

[0063] It can be seen from this that the present application can determine the speaking timing of the intelligent agent based on the speaking frequency of the intelligent agent in the asynchronous communication group, and can independently decide the speaking timing in an asynchronous communication environment without fixed communication rounds. Through dynamic scheduling prompt word adjustment, the intelligent agent can adaptively balance the roles of speaking and listening, better simulate the behavior of real users in asynchronous communication groups, and can adapt to asynchronous communication scenarios such as online team meetings; and, it can dynamically adjust the speaking behavior of the intelligent agent in combination with the real-time online members of the asynchronous communication group, so that the speaking behavior of the intelligent agent can be more adapted to the actual scenario.

[0064] like Figure 4 As shown, the embodiment of the present application discloses an intelligent agent message sending device based on an asynchronous communication group, comprising:

[0065] A speaking frequency determination module 11 is used to determine the speaking frequency of an agent in a current time period of an asynchronous communication group; the asynchronous communication group is a group related to an online team meeting;

[0066] A speech scheduling information generating module 12 is configured to generate current speech scheduling information corresponding to the agent based on a matching relationship between the speech frequency and a preset frequency threshold; the preset frequency threshold is a threshold value constructed based on the number of real-time online members of the asynchronous communication group; group members in the asynchronous communication group include real user members and / or the agent;

[0067] The message generation module 13 is used to determine whether to generate a message to be sent based on the current speech scheduling information in combination with the messages sent by each group member within the current time period, so that when the message to be sent is generated, the message to be sent is sent to the asynchronous communication group through the agent.

[0068] It can be seen from this that the present application can determine the speaking timing of the intelligent agent based on the speaking frequency of the intelligent agent in the asynchronous communication group, can independently decide the speaking timing in an asynchronous communication environment without fixed communication rounds, and can adapt to asynchronous communication scenarios such as online team meetings; and, can dynamically adjust the speaking behavior of the intelligent agent in combination with the real-time online members of the asynchronous communication group, so that the speaking behavior of the intelligent agent can be more adapted to the actual scenario.

[0069] In a specific embodiment, the speech frequency determination module 11 may include:

[0070] A message updating unit, configured to update the message corresponding to the current time period in real time according to the timestamp of the historical message corresponding to the asynchronous communication group, so as to obtain a first current message corresponding to the current time period;

[0071] a message determining unit, configured to determine a second current message sent by the agent from the first current message;

[0072] A speaking frequency determining unit is used to determine the speaking frequency of the agent in the current time period according to the message quantity relationship between the second current message and the first current message.

[0073] In a specific embodiment, the speech scheduling information generating module 12 may include:

[0074] A first speech scheduling information generating submodule is configured to generate, for the agent, current speech scheduling information indicating that the agent is allowed to speak when the speech frequency matches a preset frequency threshold;

[0075] The second speech scheduling information generating submodule is configured to generate, for the agent, current speech scheduling information indicating that the agent is not allowed to speak when the speech frequency does not match the preset frequency threshold.

[0076] In another specific embodiment, the first speech scheduling information generating submodule may include:

[0077] A first speech scheduling information generating unit is configured to determine that the speech frequency and the preset frequency threshold are in a matching relationship when the preset frequency threshold is the ratio of the target number to the number of real-time online members, and the speech frequency is less than the preset frequency threshold, and to generate current speech scheduling information indicating that speech is allowed for the agent; wherein the target number is 1.

[0078] In another specific embodiment, the first speech scheduling information generating submodule may include:

[0079] The second speech scheduling information generating unit is used to determine that the speech frequency and the preset frequency threshold are in a matching relationship when the preset frequency threshold is a threshold for controlling the speech frequency of the intelligent agent calculated based on the target weight and the number of real-time online members, and the speech frequency is less than the preset frequency threshold, and generate current speech scheduling information representing the permission of speaking for the intelligent agent; wherein the target weight includes a first target weight and a second target weight, the first target weight is the preset weight corresponding to each of the real-time online members, and the second target weight is the real-time weight corresponding to each of the real-time online members; the preset weight is a weight value pre-configured for all group members based on the member identity information of all group members in the asynchronous communication group, and the real-time weight is a weight value calculated in real time for all real-time online members based on the member identity information of all real-time online members in the current asynchronous communication group.

[0080] In a specific embodiment, the message generating module 13 may include:

[0081] a prohibition message generating unit, configured to prohibit generation of a to-be-sent message when the current speech scheduling information indicates that speaking is not permitted;

[0082] A message generation unit is configured to generate a message to be sent using a preset large language model based on a preset message generation strategy and in combination with messages sent by each of the group members within the current time period when the current speech scheduling information indicates that speaking is allowed; wherein the preset message generation strategy is a strategy constructed based on a prompt word project, a word masking operation, and a semantic repetition rate to guide the preset large language model to generate messages, and the prompt word project is a layered prompt word project that combines member roles, group communication scenarios, and historical messages.

[0083] In a specific embodiment, the message generating module 13 may include:

[0084] A message length determining unit, configured to determine the message length of the message to be sent;

[0085] a delay time setting unit, configured to set a delay time matching the message length based on a preset manual typing speed;

[0086] A message sending unit is used to send the message to be sent to the asynchronous communication group through the agent using the delay time.

[0087] Furthermore, the embodiment of the present application also discloses an electronic device, Figure 5 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram should not be considered as any limitation to the scope of application of the present application.

[0088] Figure 5 This is a schematic diagram of the structure of an electronic device 20 provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps of the method for sending messages to an agent based on an asynchronous communication group as disclosed in any of the aforementioned embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0089] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device. The communication protocol it follows is any communication protocol that can be applied to the technical solution of this application and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world. Its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0090] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or CD, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0091] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20, and can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of implementing the method for sending agent messages based on an asynchronous communication group and executed by the electronic device 20 as disclosed in any of the aforementioned embodiments, the computer program 222 can further include computer programs capable of implementing other specific tasks.

[0092] Furthermore, this application discloses a computer-readable storage medium for storing a computer program; wherein, when executed by a processor, the computer program implements the aforementioned method for sending messages to an agent based on an asynchronous communication group. The specific steps of this method can be found in the corresponding contents disclosed in the aforementioned embodiments and will not be further described here.

[0093] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from the other embodiments. Reference can be made to the descriptions of the identical or similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple, and the relevant parts can be referred to the descriptions of the methods.

[0094] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0095] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in random access memory (RAM), internal memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0096] Finally, it should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprise," "include," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus comprising a set of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus comprising the element.

[0097] The above is a detailed introduction to the technical solution provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for those skilled in the art, according to the ideas of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.< / send> < / wait> < / send>

Claims

1. A method for sending messages to an agent based on an asynchronous communication group, characterized in that: include: Determining the speaking frequency of the agent in a current time period of an asynchronous communication group; the asynchronous communication group is a group related to an online team meeting; Generating current speech scheduling information corresponding to the agent according to a matching relationship between the speech frequency and a preset frequency threshold; The preset frequency threshold is a threshold constructed based on the number of real-time online members of the asynchronous communication group; The group members in the asynchronous communication group include real user members and / or the intelligent agents; Based on the current speech scheduling information, it is determined whether to combine the messages sent by each group member in the current time period to generate a message to be sent, so that when the message to be sent is generated, the message to be sent is sent to the asynchronous communication group through the agent.

2. The method for sending messages to an agent based on an asynchronous communication group according to claim 1, characterized in that: The determining of the speaking frequency of the agent in the current time period of the asynchronous communication group includes: According to the timestamps of the historical messages corresponding to the asynchronous communication group, the messages corresponding to the current time period are updated in real time to obtain a first current message corresponding to the current time period; Determining a second current message sent by the agent from the first current message; The speaking frequency of the agent in the current time period is determined based on the message quantity relationship between the second current message and the first current message.

3. The method for sending agent messages based on asynchronous communication groups according to claim 1, characterized in that: The generating of current speech scheduling information corresponding to the agent according to the matching relationship between the speech frequency and the preset frequency threshold includes: If the speaking frequency matches the preset frequency threshold, generating current speaking scheduling information indicating that the agent is allowed to speak; If the speaking frequency does not match the preset frequency threshold, current speaking scheduling information indicating that speaking is not allowed is generated for the agent.

4. The method for sending agent messages based on asynchronous communication groups according to claim 3, characterized in that: If the speaking frequency matches the preset frequency threshold, generating current speaking scheduling information indicating that the agent is allowed to speak includes: If the preset frequency threshold is the ratio between the target number and the number of real-time online members, and the speech frequency is less than the preset frequency threshold, it is determined that the speech frequency and the preset frequency threshold are in a matching relationship, and current speech scheduling information representing the allowed speech is generated for the agent; wherein, the target number is 1.

5. The method for sending agent messages based on asynchronous communication groups according to claim 3, characterized in that: If the speaking frequency matches the preset frequency threshold, generating current speaking scheduling information indicating that the agent is allowed to speak includes: If the preset frequency threshold is a threshold for controlling the speaking frequency of the agent, calculated based on the target weight and the number of real-time online members, and the speaking frequency is less than the preset frequency threshold, then it is determined that the speaking frequency and the preset frequency threshold are in a matching relationship, and current speaking scheduling information indicating that the agent is allowed to speak is generated for the agent; Among them, the target weight includes a first target weight and a second target weight, the first target weight is a preset weight corresponding to each of the real-time online members, and the second target weight is a real-time weight corresponding to each of the real-time online members; the preset weight is a weight value pre-configured for all group members based on the member identity information of all group members in the asynchronous communication group, and the real-time weight is a weight value calculated in real time for all real-time online members based on the member identity information of all real-time online members in the current asynchronous communication group.

6. The method for sending agent messages based on asynchronous communication groups according to claim 1, characterized in that: The determining, based on the current speech scheduling information, whether to combine messages sent by each group member within the current time period to generate a message to be sent includes: If the current speech scheduling information indicates that speaking is not allowed, then the generation of the message to be sent is prohibited; If the current speech scheduling information indicates that speaking is allowed, based on a preset message generation strategy and in combination with messages sent by each of the group members in the current time period, a message to be sent is generated using a preset large language model; Among them, the preset message generation strategy is a strategy constructed based on prompt word engineering, word blocking operation and semantic repetition rate to guide the preset large language model to generate messages. The prompt word engineering is a layered prompt word engineering that combines member roles, group communication scenarios and historical messages.

7. The method for sending agent messages based on an asynchronous communication group according to any one of claims 1 to 6, characterized in that: The sending of the to-be-sent message to the asynchronous communication group by the agent includes: Determining the message length of the message to be sent; Setting a delay time matching the message length based on a preset manual typing speed; The message to be sent is sent to the asynchronous communication group using the delay time through the agent.

8. An intelligent message sending device based on an asynchronous communication group, characterized in that: include: A speaking frequency determination module is used to determine the speaking frequency of an agent in a current time period of an asynchronous communication group; the asynchronous communication group is a group related to an online team meeting; A speech scheduling information generation module, configured to generate current speech scheduling information corresponding to the agent based on a matching relationship between the speech frequency and a preset frequency threshold; The preset frequency threshold is a threshold constructed based on the number of real-time online members of the asynchronous communication group; The group members in the asynchronous communication group include real user members and / or the intelligent agents; The message generation module is used to determine whether to generate a message to be sent based on the current speech scheduling information in combination with the messages sent by each group member within the current time period, so that when the message to be sent is generated, the message to be sent is sent to the asynchronous communication group through the agent.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the agent message sending method based on an asynchronous communication group as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that Used to store a computer program, which, when executed by a processor, implements the agent message sending method based on an asynchronous communication group as described in any one of claims 1 to 7.

Images