Service providing method, transaction processing method, information display method, computing device, storage medium, and program product

Abstract

The embodiment of the invention provides a service providing method, a transaction processing method, an information display method, computing equipment, a storage medium and a program product. The server side can obtain a first message which is sent by a first user based on a session channel with a second user and aims at a target object; extracting a first bid provided by the first user based on the first message and at least one first historical message corresponding to the session channel under the condition of identifying that the first user has a bargaining intention based on the first message; comparing the first bid with the price range of the target object; generating a second message according to the comparison result; and sending the second message to the first user. According to the technical scheme provided by the embodiment of the invention, the server automatically replies the message, intelligent customer service is realized, the communication efficiency and timeliness are ensured, and the service experience of the user is improved.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computer technology, and in particular to a service providing method, a transaction processing method, an information display method, a computing device, a storage medium, and a program product. Background Art

[0002] In some online systems that provide objects for users to perform interactive behaviors, for example, online trading platforms that provide object transactions, in order to provide users with a better service experience, the online system can usually provide communication capabilities so that users and object providers can establish a conversation channel, communicate through messages, and negotiate on issues such as logistics and fees, making it easier for object providers to provide services to users.

[0003] In actual applications, especially in online trading systems that provide second-hand goods transactions in the C2C (Consumer to Consumer) model, the object provider may often be offline, resulting in a lack of communication and interaction, which affects communication efficiency and poor user service experience. Summary of the invention

[0004] The embodiments of the present application provide a service providing method, a transaction processing method, an information display method, a computing device, a storage medium and a program product to solve the technical problem of low efficiency of online system communication in the prior art.

[0005] In a first aspect, an embodiment of the present application provides a method for providing a service, including:

[0006] Acquire a first message sent by the first user to the target object based on the conversation channel with the second user;

[0007] Based on the first message, if it is identified that the first user has a bargaining intention, extracting a first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel;

[0008] comparing the first bid to a price range of the target object;

[0009] generating a second message according to the comparison result;

[0010] The second message is sent to the first user.

[0011] Optionally, the method further comprises:

[0012] When it is identified that the first user has no bargaining intention, a second message is generated based on the object description information of the target object, the first message and at least one second historical message.

[0013] Optionally, generating a second message according to the comparison result includes:

[0014] If the first bid does not exceed the price range, generating a transaction instruction, and generating a second message based on the transaction instruction;

[0015] If the first bid is higher than the upper limit of the price range, generating a rejection instruction, and generating a second message based on the rejection instruction;

[0016] When the first bid is lower than the lower limit of the price range, and the upper limit of the price is different from the lower limit of the price, selecting the first bid from the price range and generating a second message based on the first bid or generating a bid instruction, and generating a second message based on the bid instruction;

[0017] When the first bid is lower than the lower limit of the price range and the upper limit is the same as the lower limit, a rejection instruction is generated and a second message is generated based on the rejection instruction, or a bid instruction is generated and a second message is generated based on the bid instruction.

[0018] Optionally, after the second message is generated, the method further includes:

[0019] In a case where the first bid is lower than a lower price limit of the price range and a first offer is selected from the price range, an upper price limit of the price range is updated using the first offer.

[0020] Optionally, after the second message is generated, the method further includes:

[0021] In the case where the first bid is lower than the lower limit of the price range, the highest bid is selected from the historical bids of the first user to update the lower limit of the price range.

[0022] Optionally, the method further comprises:

[0023] In the case where the first bid is not obtained, searching from the historical message record of the conversation channel whether there is a second bid;

[0024] Based on the search result, a second message is generated.

[0025] Optionally, generating a second message based on the search result includes:

[0026] If no second quotation exists, selecting a second quotation from the price range, and generating a second message based on the second quotation;

[0027] In the case where there is a second bid, determining whether there is a second bid of the first user in the historical message record,

[0028] If so, updating the price range based on the second quotation and the second bid, and selecting a third quotation from the price range and generating a second message based on the third quotation, or generating a bid instruction and generating a second message based on the bid instruction;

[0029] If not, the price range is updated based on the second quotation, and a third quotation is selected from the price range and a second message is generated based on the third quotation, or a bid instruction is generated and a second message is generated based on the bid instruction.

[0030] Optionally, when the first user is identified to have a bargaining intention based on the first message, extracting a first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel includes:

[0031] When the customer service robot identifies the first user's intention to bargain based on the first message, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel;

[0032] Generating a second message according to the comparison result includes:

[0033] Based on the comparison result, a second message is generated using the customer service robot.

[0034] Optionally, when the first user is identified to have a bargaining intention based on the first message, extracting a first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel includes:

[0035] Inputting the first message into an intention recognition model to identify whether the first user has a bargaining intention;

[0036] In the case that the first user has bargaining intention, a first bid provided by the first user is extracted based on the first message and at least one first historical message corresponding to the conversation channel by using a first large model.

[0037] Optionally, the intent recognition model is trained in the following manner:

[0038] The intention recognition model is trained using the second sample data and the training label of the second sample data; the training label of the second sample data includes whether the second sample data has bargaining intention.

[0039] Optionally, the first large model is trained in the following manner:

[0040] The first large model is trained using the third sample data and the training labels of the third sample data; the training labels of the third sample data include the price data included in the third sample data.

[0041] Optionally, when it is identified that the first user has no bargaining intention, generating the second message based on the object description information of the target object, the first message and at least one second historical message includes:

[0042] When it is identified that the first user has no bargaining intention, a second message is generated using a second large model based on the object description information of the target object, the first message and at least one second historical message.

[0043] Optionally, generating a second message based on the transaction instruction includes: generating a second message based on the transaction instruction using a second large model;

[0044] The generating a second message based on the rejection instruction includes: generating a second message based on the rejection instruction using the second large model;

[0045] The generating a second message based on the first quotation includes: generating a second message based on the first quotation using the second large model

[0046] The generating a second message based on the bidding instruction includes: generating a second message based on the bidding instruction using the second large model.

[0047] Optionally, the second largest model is trained in the following manner:

[0048] The second large model is trained using the fourth sample data and the training labels of the fourth sample data; the training labels of the third sample data include the response data corresponding to the fourth sample data.

[0049] Optionally, before identifying the first message, the method further includes:

[0050] Using the detection model, detecting whether the first message is an abnormal message;

[0051] If not, performing an identification operation on the first message;

[0052] If so, a second message is generated based on the predetermined message.

[0053] Optionally, the method further comprises:

[0054] When the first bid does not exceed the price range, sending transaction prompt information for the target object to the first user;

[0055] In response to the first user's order request for the target object, generating a transaction order according to the first bid;

[0056] A success prompt message is sent to the second user.

[0057] In a second aspect, an embodiment of the present application provides a transaction processing method, including:

[0058] Acquire a first message sent by the first user to the target object based on the conversation channel with the second user;

[0059] Based on the first message, if it is identified that the first user has a bargaining intention, extracting a first bid of the first user based on the first message and at least one first historical message corresponding to the conversation channel;

[0060] comparing the first bid with a price range corresponding to the target object;

[0061] When the comparison result meets the transaction requirements, a transaction reminder message for the target object is sent to the first user.

[0062] In a third aspect, an information display method is provided in an embodiment of the present application, comprising:

[0063] Providing a conversation interface between the first user and the second user;

[0064] In response to a message sending operation triggered by the first user on the conversation interface, determining a first message provided by the first user and directed to a target object;

[0065] Sending the first message to the server, so that the server can identify, based on the first message, that the first user has a bargaining intention, extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel between the first user and the second user; comparing the first bid with the price range of the target object, and generating a second message according to the comparison result;

[0066] The second message is obtained, and the second message is displayed in the conversation interface.

[0067] Optionally, the method further comprises:

[0068] Acquire transaction prompt information sent by the server; the transaction prompt information is generated when the comparison result meets the transaction requirements.

[0069] The transaction prompt information is displayed in the conversation interface.

[0070] In a fourth aspect, a computing device is provided in an embodiment of the present application, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the service provision method as described in the first aspect above, the transaction processing method as described in the second aspect above, or the information display method as described in the third aspect above.

[0071] In the fifth aspect, a computer storage medium is provided in an embodiment of the present application, storing a computer program, which, when executed by a computer, implements the service providing method as described in the first aspect above, the transaction processing method as described in the second aspect above, or the information display method as described in the third aspect above.

[0072] In the sixth aspect, a computer program product is provided in an embodiment of the present application, including a computer program / instructions, which, when executed by a computer, implements the service provision method as described in the first aspect above, the transaction processing method as described in the second aspect above, or the information display method as described in the third aspect above.

[0073] In the embodiment of the present application, the server can obtain the first message sent by the first user to the target object based on the conversation channel with the second user; based on the first message, identify the first user's bargaining intention, and extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel; compare the first bid with the price range of the target object; generate a second message based on the comparison result; and send the second message to the first user. In the embodiment of the present application, the server automatically replies to the message to realize intelligent customer service, ensure communication efficiency and timeliness, and improve the user's service experience.

[0074] These and other aspects of the present application will become more clearly understood in the description of the following embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0076] Figure 1 A flowchart of an embodiment of a service providing method provided by the present application is shown;

[0077] Figure 2 A flow chart showing an embodiment of a transaction processing method provided by the present application is shown;

[0078] Figure 3 A flowchart of an embodiment of an information display method provided by the present application is shown;

[0079] Figure 4 A schematic diagram of the interface display of a first user terminal in a practical application of an embodiment of the present application is shown;

[0080] Figure 5 A schematic diagram of scene interaction in an actual application of an embodiment of the present application is shown;

[0081] Figure 6 A schematic diagram of a bargaining process in a practical application of an embodiment of the present application is shown;

[0082] Figure 7 A schematic diagram showing the structure of an embodiment of a service providing device provided by the present application is shown;

[0083] Figure 8 A schematic diagram showing the structure of an embodiment of a transaction processing device provided by the present application is shown;

[0084] Fig. 9 A structural schematic diagram of another embodiment of an information display device provided by the present application is shown;

[0085] Fig.10 A schematic diagram of the structure of an embodiment of a computing device provided by the present application is shown. DETAILED DESCRIPTION

[0086] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0087] In some of the processes described in the specification and claims of this application and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish between different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0088] The technical solution of the embodiment of the present application can be applied to an online system that provides objects for users to perform interactive behaviors, such as an online trading system that provides second-hand goods transactions. In these online systems, the object provider, that is, the seller, will set the price of the object, and the user, that is, the buyer, can pay resource data of the same price, such as currency, to interact with the object. With the help of the conversation channel provided by the online system, the seller can provide customer service, the buyer will consult the seller for relevant information about the object, and will also negotiate with the seller on the price, etc. The buyer will often bargain, that is, may provide a bid lower than the price of the object, and the seller will also re-quote based on the buyer's bid, etc., and finally the buyer and the seller reach an agreement to facilitate the transaction, all of which require efficient and timely communication.

[0089] However, since the object provider may often be offline, communication cannot be carried out, and thus interaction cannot be facilitated, which affects communication efficiency, poor user service experience, and low object conversion rate.

[0090] In order to solve the problem of communication efficiency, the inventor proposed the technical solution of the present application after a series of studies. In the embodiment of the present application, the server can obtain the first message sent by the first user to the target object based on the conversation channel with the second user; based on the first message, identify that the first user has the intention to bargain, and extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel; compare the first bid with the price range of the target object; generate a second message according to the comparison result; and send the second message to the first user. In the embodiment of the present application, the server automatically replies to messages to realize intelligent customer service, ensure communication efficiency and timeliness, and through bargaining intention identification, bid extraction and price comparison, etc., it replaces the second user to bargain with the first user, avoiding the interruption of bargaining communication when the second user is offline, further ensuring communication continuity, improving communication efficiency, helping to improve the transaction rate of the target object, and improving the user's service experience.

[0091] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0092] It should be noted that, although the first user and the second user are described in this article, those skilled in the art can understand that the first user and the second user actually correspond to different user accounts in the server, which can be obtained by pre-registration by the user; the interactive operations between the server and different users are based on the user account, and the corresponding data received or sent by the server to the user is also based on the user account, and the corresponding data is actually received or sent by the user end corresponding to the user account to the server. Users can communicate with each other through user accounts.

[0093] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0094] It should be noted that the technical solution of the embodiment of the present application is applicable to a network virtual environment. The users described generally refer to "virtual users". Real users can register a user account in the server through registration to obtain a user identity in the network environment.

[0095] The implementation details of the technical solution of the embodiment of the present application are elaborated in detail below.

[0096] Figure 1 A flowchart of an embodiment of a service providing method provided by the present application. The technical solution of this embodiment can be executed by a server, and the server is deployed in an online system. In actual applications, the online system is usually composed of a first user terminal, a second user terminal and a server, and the first user terminal and the second user terminal are connected to the server through a network. The network provides a medium for the communication link between the first user terminal and the second user terminal and the server. The network can include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0097] The first user terminal can be oriented to the first user, that is, the consumer user or buyer in actual application, and the second user terminal can be oriented to the second user, that is, the object provider or seller. The second user can publish objects through the second user terminal, and the first user can perform interactive behaviors such as object consumption through the first user terminal.

[0098] The method may include the following steps:

[0099] 101: Obtain a first message sent by a first user to a target object based on a conversation channel with a second user.

[0100] The conversation channel may be established between the first user terminal and the second user terminal based on the server terminal. The first user may interact with the second user through the first user terminal and the second user may interact with the first user through the second user terminal based on the conversation channel.

[0101] The first message may be sent by a first user via a first user terminal.

[0102] For example, in an online transaction scenario, the first user may be the buyer of the target object, and the second user may be the object provider of the target object, that is, the seller. The target object may be a target commodity, and the conversation channel between the first user and the second user may be provided by an online transaction system. In an actual application, the online transaction system may be an online transaction system that provides second-hand commodity transactions, and the target object may be a target second-hand commodity.

[0103] 102: Based on the first message, if it is identified that the first user has a bargaining intention, extract a first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel.

[0104] In this embodiment, after receiving the first message, the server can first perform intent recognition based on the first message regardless of whether the second user is offline. Of course, step 102 can also be performed when the second user is offline. In addition, as another optional method, step 102 can also be performed when the second user is in a managed state. The second user being in a managed state can be set when receiving a managed request from the second user.

[0105] Among them, whether the first user has the intention to bargain can be determined based on whether the first message includes price information, and it can also be identified by using a machine learning model, etc., which will be introduced in detail in the corresponding embodiments below.

[0106] The at least one first historical message may be a historical message sent by the first user, and the first message may use the latest price information involved in the at least one first historical message as the first bid.

[0107] In addition, at least one first historical information may include a message sent by the first user, and may also include a historical message sent by the second user and / or a historical message generated by the server.

[0108] The first bid may also be extracted using a machine learning model, which will be described in detail in the corresponding embodiments below.

[0109] 103: Compare the first bid with the price range of the target object.

[0110] The price range may be composed of an upper price limit and a lower price limit corresponding to the target object. For example, the price of the target object published by the second user may be used as the upper price limit. In addition, the second user may also pre-set a minimum price and save it by the server. The minimum price may be provided by the second user when publishing the target object or when sending a hosting request. Of course, the server may also provide corresponding setting prompt information in the user interface of the user end to prompt the second user to set the minimum price. In a possible implementation, the minimum price of the target object may be used as the lower price limit of the price range.

[0111] Of course, the price range will also be updated with the latest bid or latest quotation involved in the bargaining process. The latest quotation may be provided by the second user, or it may be generated by the server, so that the latest quotation can be used as the price cap, etc.

[0112] 104: Generate a second message according to the comparison result.

[0113] The comparison result of comparing the first bid with the price range of the target object may include, for example, that the first bid is lower than the lower limit of the price range, that the first bid is within the price range, or that the first bid is higher than the upper limit of the price range. Second messages with different contents may be generated according to different comparison results. For example, different comparison results represent different transaction probabilities. If the first bid is within the price range, it indicates that the first bid meets the expectations of the second user, the target object can be traded, and the second message may include transaction information, etc. If the first bid exceeds the price range, the target object cannot be traded, and the second message may include a rejection message, etc.

[0114] Of course, the second messages corresponding to different comparison results can be preset. In addition, the second messages corresponding to different comparison results can also be generated using a machine learning model, which will be described in detail in the corresponding embodiments below.

[0115] 105: Send a second message to the first user.

[0116] In this embodiment, the server can obtain a first message sent by the first user to the target object based on the conversation channel with the second user; based on the first message, identify that the first user has the intention to bargain, and extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel; compare the first bid with the price range of the target object; generate a second message according to the comparison result; and send the second message to the first user. In the embodiment of the present application, the server automatically replies to messages to realize intelligent customer service, ensure communication efficiency and timeliness, and through bargaining intention identification, bid extraction, and price comparison, etc., the server replaces the second user to bargain with the first user, avoiding the interruption of bargaining communication when the second user is offline, further ensuring communication continuity, improving communication efficiency, helping to improve the transaction rate of the target object, and improving the user's service experience.

[0117] In some embodiments, obtaining a first message sent by a first user to a target object may include: establishing a conversation channel between the first user and a second user in response to a service request of the first user; and obtaining a first message sent by the first user to the target object based on the conversation channel.

[0118] In some embodiments, the method may further include: generating a second message based on object description information of the target object, the first message, and at least one second historical message when it is identified that the user has no bargaining intention.

[0119] The object description information may be provided by the object provider, such as detailed introduction information in the object details page, etc., including, for example, attribute parameters, brand, purpose and / or design style of the target object.

[0120] Among them, the user's needs and problems can be located based on the first message of the first user and at least one second historical message. Since the user has no intention of bargaining, the first user's message may be a question about the target object, etc., hoping to obtain detailed information about the target object. Therefore, the second message generated based on the object description information of the target object can answer the user's questions about the target object.

[0121] Among them, when the first message provided by the first user does not raise questions for the target object, etc., and the server has difficulty in answering the first message, the generated second message can be a predetermined message with friendly content.

[0122] In some embodiments, generating the second message according to the comparison result may include:

[0123] If the first bid price does not exceed the price range, a transaction instruction is generated, and based on the transaction instruction, a second message is generated;

[0124] generating a rejection instruction if the first bid is higher than the upper price limit of the price range, and generating a second message based on the rejection instruction;

[0125] When the first bid is lower than the lower limit of the price range, and the upper limit of the price is different from the lower limit of the price, selecting the first bid from the price range and generating a second message based on the first bid or generating a bid instruction and generating a second message based on the bid instruction;

[0126] When the first bid is lower than the lower limit of the price range and the upper limit is the same as the lower limit, a rejection instruction is generated and a second message is generated based on the rejection instruction, or a bid instruction is generated and a second message is generated based on the bid instruction.

[0127] Among them, the first bid does not exceed the price range, indicating that the first bid is reasonable, so a transaction instruction can be generated directly.

[0128] Among them, the first bid is higher than the upper limit of the price range, indicating that the first bid is unreasonable, so a rejection instruction can be directly generated.

[0129] Among them, when the first bid is lower than the lower limit of the price range and the upper price limit is different from the lower price limit, optionally, it is possible to determine according to a first preset probability whether to select the first quotation from the price range and generate a second message based on the first quotation, or to generate a bidding instruction and generate a second message based on the bidding instruction.

[0130] For example, an 80% probability is set to execute the operation of selecting a first quotation from a price range and generating a second message based on the first quotation, and a 20% probability is set to execute the operation of generating a bid instruction and generating a second message based on the bid instruction.

[0131] Optionally, selecting the first quotation from the price range may be to obtain the first quotation by random sampling in the price range according to a probability density function such as a Gaussian distribution.

[0132] Among them, when the first bid is lower than the lower limit of the price range and the upper limit of the price is the same as the lower limit of the price, optionally, a rejection instruction can be generated according to a second preset probability and a second message can be generated based on the rejection instruction, or a bidding instruction can be generated and a second message can be generated based on the bidding instruction.

[0133] For example, a probability of 20% is set to execute the operation of generating a rejection instruction, and a probability of 80% is set to execute the operation of generating a bid instruction and generating a second message based on the bid instruction.

[0134] Optionally, the second message generated based on the bidding instruction may include content guiding the first user to increase the bid.

[0135] In some embodiments, after the second message is generated, the method may further include: if the first bid is lower than a lower price limit of the price range and the first offer is selected from the price range, updating an upper price limit of the price range using the first offer.

[0136] The first quotation selected from the price range is usually lower than the original price cap. Updating the price cap using the first quotation can lower the price cap and make the price range more reasonable.

[0137] In some embodiments, after the second message is generated, the method may further include:

[0138] In the case that the first bid is lower than the lower limit of the price range, the highest bid is selected from the historical bids of the first user to update the lower limit of the price range.

[0139] By setting the first user's highest historical bid as the new price floor, the price setting can be made more reasonable, avoiding difficulty in closing a deal due to the initial price floor being too high, and increasing the probability of closing a deal.

[0140] In some embodiments, if the first bid is not extracted, the method may further include:

[0141] In the case that the first bid is not obtained, whether the second bid exists is searched from the historical message record of the conversation channel; and a second message is generated based on the search result.

[0142] The second quotation may be proposed by the server or by the second user. The second quotation may be selected by the server from a price range during a historical bargaining process.

[0143] In some embodiments, when the first bid is not extracted, generating the second message based on the search result may include:

[0144] If no second quotation exists, selecting a second quotation from the price range, and generating a second message based on the second quotation;

[0145] If there is a second bid, determine whether there is a second bid of the first user in the historical message record.

[0146] If so, updating the price range based on the second quotation and the second bid, and selecting a third quotation from the price range and generating a second message based on the third quotation, or generating a bid instruction and generating a second message based on the bid instruction;

[0147] If not, the price range is updated based on the second quotation, and a third quotation is selected from the price range and a second message is generated based on the third quotation, or a bid instruction is generated and a second message is generated based on the bid instruction.

[0148] Optionally, in the case where there is no second quotation, selecting the second quotation from the price range may be to obtain the second quotation by random sampling in the price range according to a probability density function such as a Gaussian distribution.

[0149] Optionally, when there is a second quotation and there is a second bid of the first user in the historical message record, the price range can be updated based on the second quotation and the second bid according to a third preset probability, and a third quotation can be selected from the price range and a second message can be generated based on the third quotation, or a bidding instruction can be generated and a second message can be generated based on the bidding instruction.

[0150] Among them, the judgment that there is a second offer and there is a second offer of the first user in the historical message record is based on the fact that the first offer of the first user has not been extracted. The first offer of the first user is extracted based on the first message and at least one first historical message, and the second offer of the first user is judged based on the historical message record. In the case where the number of messages included in the historical message record is greater than the number of messages in the at least one first historical message, there may be a situation where the second offer of the first user exists in the historical message record, and the first offer of the first user cannot be extracted from the first message and at least one first historical message.

[0151] For example, the operations of updating the price range based on the second quotation and the second bid, selecting the third quotation from the price range and generating the second message based on the third quotation can be performed with an 80% probability, and the operations of generating a bidding instruction and generating the second message based on the bidding instruction can be performed with a 20% probability.

[0152] Among them, updating the price range based on the second quotation and the second bid, and selecting the third quotation from the price range can be to use the second quotation to update the price upper limit of the price range, the second bid to update the price lower limit of the price range, and randomly sample the third quotation in the updated price range according to probability density functions such as Gaussian distribution.

[0153] Optionally, when there is a second quotation and there is no second bid of the first user in the historical message record, the price range can be updated based on the second quotation according to a fourth preset probability, and a third quotation can be selected from the price range and a second message can be generated based on the third quotation, or a bidding instruction can be generated and a second message can be generated based on the bidding instruction.

[0154] For example, the operations of updating the price range based on the second quotation, selecting a third quotation from the price range and generating a second message based on the third quotation may be performed with an 80% probability, and the operations of generating a bidding instruction and generating a second message based on the bidding instruction may be performed with a 20% probability.

[0155] Among them, updating the price range based on the second quotation and selecting the third quotation from the price range can be performed by using the second quotation to update the price upper limit of the price range, and randomly sampling the third quotation within the updated price range according to a probability density function such as Gaussian distribution.

[0156] In some embodiments, before identifying the first message, the method may further include: using a detection model to detect whether the first message is an abnormal message; if not, performing an identification operation on the first message; if so, generating a second message based on a predetermined message.

[0157] The scheduled message may be, for example, a content-friendly scheduled message. By replying to the content-friendly scheduled message, the negative emotions of the customer can be alleviated and potential conflict escalation can be reduced.

[0158] The abnormal message may refer to a message including insulting, hateful, threatening and / or pornographic content, for example.

[0159] The detection model may be obtained by training, for example, in the following manner: using the first sample data and the training label of the first sample data to train the intent recognition model; the training label of the first sample data includes whether the first sample data is an abnormal message.

[0160] The first sample data may come from, for example, a question-answering knowledge base, an open source conversation dataset, and / or real conversation data from an e-commerce platform.

[0161] Compared with a large language model with a large pre-training data scale and a large number of parameters, the detection model can be trained by a small or medium-sized model or a lightweight model with a small pre-training data scale and a small number of parameters, so that the detection operation can be completed with less computing resources. For example, the detection model can be trained based on an autoencoder model such as BERT (Bidirectional Encoder Representations from Transformers, a bidirectional encoder model based on Transformers).

[0162] This embodiment uses the detection model to identify abnormal messages and promptly reply to the scheduled message, which can reduce potential customer complaints and disputes, reduce corporate risks, and improve communication efficiency. After identifying the abnormal message, subsequent operations such as the first message identification will not be performed, which can save resources.

[0163] In some embodiments, based on the first message, when the first user is identified to have a bargaining intention, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel may include: using a customer service robot to identify, based on the first message, when the first user is identified to have a bargaining intention, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel;

[0164] Generating a second message according to the comparison result may include: generating the second message using a customer service robot according to the comparison result.

[0165] Among them, the customer service robot has the ability of machine learning and has a high ability to understand context, which can extract key content more accurately. Using the customer service robot to extract the user's first bid is more flexible, accurate and efficient than extracting information using preset rules. And using the customer service robot to generate the second message can reduce mechanicalness and stiffness, and generate more fluent and natural content, compared with generating the second message using preset rules.

[0166] Optionally, the customer service robot may integrate a detection model, an intention recognition model, a first large model, and a second large model. Among them, the detection model of the customer service robot may be used to detect whether the first message is an abnormal message. Using the customer service robot to identify whether the first user has a bargaining intention based on the first message, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel may be using the intention recognition model of the customer service robot to identify whether the first user has a bargaining intention based on the first message, and when the first user has a bargaining intention, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel using the first large model of the customer service robot. Generating the second message using the customer service robot may be generating the second message using the second large model of the customer service robot. Specific explanations may be described below.

[0167] In some embodiments, when it is identified based on the first message that the first user has the intention to bargain, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel may include: inputting the first message into an intention recognition model to identify whether the first user has the intention to bargain; when the first user has the intention to bargain, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel using the first large model.

[0168] Optionally, the intention recognition model can be trained in the following manner: the intention recognition model is trained using the second sample data and the training label of the second sample data; the training label of the first sample data includes whether the second sample data has bargaining intention.

[0169] The second sample data may come from, for example, a question-and-answer knowledge base, an open source conversation dataset, and / or real conversation data from an e-commerce platform.

[0170] Compared with large language models with large pre-training data scale and parameters, the intent recognition model can be a small or medium-sized model or a lightweight model with small pre-training data scale and parameters, so that the recognition operation can be completed with less computing resources. For example, the intent recognition model can be trained based on an autoencoder model such as BERT.

[0171] The first large model may be obtained by training in the following manner: the first large model is trained using the third sample data and the training labels of the third sample data; the training labels of the third sample data include the price data included in the third sample data.

[0172] The third sample data may come from, for example, a question-and-answer knowledge base, an open source conversation dataset, and / or real conversation data from an e-commerce platform.

[0173] Among them, using the third sample data and the training labels of the second sample data, training the first large model, for example, can be using the third sample data and the training labels of the third sample data to generate prompt instructions, inputting the prompt instructions into the first large language model, and training the first large language model. The first large model can be a pre-trained large language model (English: Large Language Model, abbreviated as: LLM). Since the information to be processed for extracting the first bid is more complex and diverse, more computing resources are required, and higher precision is required, a large language model with a large pre-training data scale and a large number of parameters can be selected, and the first large model can be obtained by fine-tuning it through the third sample data. For example, the fine-tuning method such as SFT (Supervised Fine-Tuning) can be used to fine-tune the first large model. Since the large model may be prone to hallucination problems and the generated text does not meet the format specifications, by fine-tuning the first large model using the third sample data, the first large model obtained can learn more information about the text format, thereby generating text that is more in line with the format specifications. It can also better adapt to specific tasks such as extracting the first bid from the message, and can reduce the false alarm rate of the large model when extracting numbers, and improve accuracy.

[0174] Among them, the large language model refers to a large deep learning model based on massive data and computing resources, which is used for natural language processing tasks such as language generation, text classification, machine translation, etc. It is an AI (Artificial Intelligence) model. The large language model is trained by a deep learning algorithm and can learn a large amount of text data to generate natural language answers that humans can understand. When generating answers, the large language model will obtain the most likely answer through calculation and reasoning based on the input question and context information, and output it. At the same time, it can also continuously optimize and improve its answering ability based on user feedback. In practical applications, the first large model can be, for example, a large language model such as GPT (Generative Pre-trained Transformer, a generative pre-training model), LLaMA (Large Language Model Meta AI, a large language model) or ChatGLM (Chat Generative Language Model, a chat generative language model).

[0175] In some embodiments, the second message may be generated using the second large model. When it is identified that the first user has no bargaining intention, generating the second message based on the object description information of the target object, the first message, and at least one second historical message may include: when it is identified that the first user has no bargaining intention, generating the second message based on the object description information of the target object, the first message, and at least one second historical message using the second large model.

[0176] In some embodiments, generating the second message based on the transaction instruction may include: generating the second message based on the transaction instruction using a second large model;

[0177] Based on the rejection instruction, generating the second message may include: generating the second message based on the rejection instruction using the second large model;

[0178] Generating the second message based on the first offer may include: generating the second message based on the first offer using a second large model;

[0179] Generating the second message based on the bidding instruction may include generating the second message based on the bidding instruction using a second large model.

[0180] Optionally, the second largest model can be obtained by training in the following manner: the second largest model is trained using the fourth sample data and the training label of the fourth sample data; the training label of the fourth sample data includes the response data corresponding to the fourth sample data.

[0181] The fourth sample data may come from, for example, a question-and-answer knowledge base, an open source conversation data set, and / or real conversation data from an e-commerce platform.

[0182] Among them, using the fourth sample data and the training label of the fourth sample data, training the second largest model, for example, can be to generate a prompt instruction using the fourth sample data and the training label of the fourth sample data, input the prompt instruction into the second largest model, and fine-tune the second largest model. The second largest model can be a pre-trained large language model. Since the task complexity of generating the second message is high, the model needs to have strong language understanding and generation capabilities. Therefore, a large language model with a large pre-training data scale, a large number of parameters, and a strong language understanding and generation capability can be used as the second largest model. For example, the fine-tuning method such as SFT can be used to fine-tune the second largest model. Among them, the second largest model can be the same as or different from the first largest model. By using the response data corresponding to the fourth sample data as a label to train the second largest model, the language understanding ability of the second largest model in the dialogue scenario can be improved, so that the second largest model can generate a more coherent and natural response.

[0183] The second message is generated based on the transaction instruction using the second largest model, which may be generated by inputting the transaction instruction into the second largest model. The transaction instruction may be preset, for example, the transaction instruction may specifically be "please help me generate a message for prompting the transaction".

[0184] The above-mentioned use of the second largest model to generate the second message based on the rejection instruction may be to input the rejection instruction into the second largest model to generate the second message. The rejection instruction may be, for example, pre-set, for example, the rejection instruction may specifically be "please help me generate a friendly rejection prompt message to remind the user to pay attention to the price". Therefore, the second message generated based on the rejection instruction may include content that kindly reminds the user to pay attention to the price.

[0185] The above-mentioned use of the second large model to generate the second message based on the first quotation may be to input the first quotation instruction generated based on the first quotation into the second large model to generate the second message. The first quotation instruction may be generated by splicing the first prompt template and the first quotation. For example, the first prompt template may be "Please help me generate a message for making a new quotation, the new quotation is: XX", and the first quotation may be spliced ​​at the end "XX" of the first prompt template, thereby generating the first quotation instruction. Of course, the form and content of the first prompt template are not limited to this.

[0186] The above-mentioned use of the second largest model to generate the second message based on the bid instruction may be to input the bid instruction into the second largest model and use the information output by the second largest model as the second message. The bid instruction may be, for example, spliced ​​based on the second prompt template, the first message and at least one third historical message, wherein the at least one third historical message may be the same as or have overlapping parts with at least one first historical message, and at least one third historical message may be the same as or have overlapping parts with at least one second historical message. For example, the second prompt template may be "Please help me generate a message to guide users to increase bids based on the message record, the message record is: XXXX", the first message and at least one third historical message may be spliced ​​at the end "XXXX" of the second prompt template, thereby generating a bid instruction. Of course, the form and content of the second prompt template are not limited thereto. The second message generated by the bid instruction may include content to guide the first user to increase the new bid.

[0187] Among them, the instructions (prompt) involved in this application are a form of input used to prompt or guide the large model to give an expected output. It is used to indicate what action the large model should take or what output it should generate when performing a specific task. An instruction is a natural language input, similar to a command or instruction, to let the large model know what it needs to do. It can be a question, a sentence, or a paragraph, which is used to provide the large model with context and parameter information of the input information, prompting or guiding the large model to give an expected output.

[0188] In addition, in some embodiments, when the first bid is not extracted and no second quotation is found in the historical message records of the conversation channel, the second quotation is selected from the price range, and the second message is generated based on the second quotation. This can be by using a second large model to generate the second message based on the second quotation.

[0189] In the above case where the first bid is not extracted and no second quotation exists in the historical message records of the conversation channel, the third quotation is selected from the updated price range, and the second message is generated based on the third quotation. The second message can be generated based on the third quotation using the second largest model.

[0190] In this application, using a detection model to detect whether the first message is an abnormal message, and using an intent recognition model to identify whether the first user has a bargaining intention are relatively simple discrimination tasks. Therefore, the detection model and the intent recognition model can be obtained by training with small and medium-sized models or lightweight models. Using the first large model to extract the first bid is a relatively complex extraction task, and using the second large model to generate the second message is a relatively complex generation reply task. The large language model has powerful language understanding and generation capabilities and higher accuracy. Therefore, the first large model and the second large model can be obtained by training with a large language model. Among them, compared with small and medium-sized models and lightweight models, large language models are more suitable for processing complex tasks and require more computing resources. Choosing different types of models for different operations can better balance task requirements, resource constraints, and performance requirements.

[0191] In some embodiments, the method may further include: sending a transaction prompt message for the target object to the first user when the first bid does not exceed the price range; generating a transaction order according to the first bid in response to the first user's order request for the target object; and sending a success prompt message to the second user.

[0192] The transaction prompt information may prompt the first user to place an order, for example, the transaction prompt information may be in the form of a pop-up window, which may include an order control, and may generate a transaction order according to the first bid in response to an order request generated by the first user's triggering operation on the order control. The transaction prompt information may be different from the second message generated based on the transaction instruction, which may be just a text message to prompt the first user to complete the transaction, for example, "OK, deal".

[0193] The success prompt information may include, for example, information related to the transaction order. Optionally, the success prompt information may be sent to the second user by an outbound call, for example, the success prompt information may be in the form of a text message. Optionally, the success prompt information may be displayed on the conversation interface of the second user terminal in the form of a message. Optionally, the success prompt information may be sent to the second user in the form of a notification, for example, the success prompt information may be in the form of a pop-up window.

[0194] In some embodiments, based on the first message, when it is identified that the first user has the intention to bargain, extracting the first bid provided by the first user based on the first message and at least one first historical message may include: when the second user is in an offline state or a hosted state, based on the first message, when it is identified that the first user has the intention to bargain, extracting the first bid provided by the first user based on the first message and at least one first historical message.

[0195] Optionally, when the second user is online, the message record between the first user and the customer service robot can be synchronized to the second user.

[0196] Optionally, the customer service robot can be paused when the second user is online.

[0197] That is, the customer service robot is enabled only when the second user is offline or in a hosted state, and is paused when the second user is online. This ensures that the second user can communicate with the first user even when not online, avoiding communication interruption due to the second user being offline and ensuring the continuity of the interaction.

[0198] Figure 2 This is a flow chart of an embodiment of a transaction processing method provided by the present application. The technical solution of this embodiment can be executed by the server. The method may include the following steps:

[0199] 201: Obtain a first message sent by a first user to a target object based on a conversation channel with a second user.

[0200] 202: Based on the first message, if it is identified that the first user has a bargaining intention, extract a first bid of the first user based on the first message and at least one first historical message corresponding to the conversation channel.

[0201] 203: Compare the first bid with the price range corresponding to the target object.

[0202] 204: When the comparison result meets the transaction requirement, a transaction prompt message for the target object is sent to the first user.

[0203] The comparison result meeting the transaction requirement may be, for example, that the first bid is within the price range.

[0204] The transaction prompt information may prompt the first user to place an order, for example, the transaction prompt information may be in the form of a pop-up window, which may include an order control, and may generate a transaction order according to the first bid in response to an order request generated by the first user's triggering operation on the order control. The transaction prompt information may be different from the second message generated based on the transaction instruction, which may be just a text message to prompt the first user to complete the transaction, for example, "OK, deal".

[0205] In this embodiment, the server can obtain a first message sent by the first user to the target object based on the conversation channel with the second user; based on the first message, identify that the first user has the intention to bargain, and extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel; compare the first bid with the price range of the target object; and send a transaction prompt message for the target object to the first user when the comparison result meets the transaction requirements. In the embodiment of the present application, the server replaces the second user to bargain with the first user through bargaining intention identification, bid extraction, price comparison, and sending of transaction prompt messages, thereby avoiding interruption of bargaining communication when the second user is offline, further ensuring communication continuity, improving communication efficiency, helping to increase the transaction rate of the target object, and improving the user's service experience.

[0206] In some embodiments, when it is identified based on the first message that the first user has the intention to bargain, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel may include: inputting the first message into an intention recognition model to identify whether the first user has the intention to bargain; when the first user has the intention to bargain, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel using the first large model.

[0207] Optionally, the intention recognition model can be trained in the following manner: the intention recognition model is trained using the second sample data and the training label of the second sample data; the training label of the first sample data includes whether the second sample data has bargaining intention.

[0208] Optionally, the first large model can be obtained by training in the following manner: the first large model is trained using the third sample data and the training labels of the third sample data; the training labels of the third sample data include the price data included in the third sample data.

[0209] The specific implementation of using the intention recognition model to identify whether the user has bargaining intention and using the first model to extract the first bid can be found in detail. Figure 1 As described in the illustrated embodiment, details will not be repeated here.

[0210] In some embodiments, the method may further include: sending a transaction reminder message for the target object to the first user if the first bid does not exceed the price range; generating a transaction order according to the first bid in response to the first user's order request for the target object; and sending a success reminder message to the second user. Figure 1 As described in the illustrated embodiment, details will not be repeated here.

[0211] Figure 3This is a flowchart of an embodiment of an information display method provided by the present application. The technical solution of this embodiment can be executed by a first user terminal. The method may include the following steps:

[0212] 301: Provide a conversation interface between the first user and the second user.

[0213] 302: In response to a message sending operation triggered by a first user in a conversation interface, determining a first message provided by the first user and directed to a target object.

[0214] 303: Send a first message to the server, so that the server can identify, based on the first message, that the first user has bargaining intention, and extract a first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel; compare the first bid with the price range of the target object; and generate a second message based on the comparison result.

[0215] 304: Obtain the second message, and display the second message in the conversation interface.

[0216] In this embodiment, the first user terminal can provide a conversation interface between the first user and the second user. In response to the message sending operation triggered by the first user in the conversation interface, the first message provided by the first user for the target object is determined, and the first message is sent to the server. The server can identify the first user's bargaining intention based on the first message, and extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel between the first user and the second user; compare the first bid with the price range of the target object; generate a second message according to the comparison result; obtain the second message, and display the second message in the conversation interface. In the embodiment of the present application, the message reply automatically performed by the server is displayed in the conversation interface to realize intelligent customer service, ensure communication efficiency and timeliness, and through bargaining intention identification, bid extraction and price comparison, etc., the second user and the first user are bargained on behalf of each other, avoiding the interruption of bargaining communication when the second user is offline, further ensuring communication continuity, improving communication efficiency, helping to improve the transaction rate of the target object, and improving the user's service experience.

[0217] In some embodiments, the method may further include: obtaining transaction prompt information sent by the server; and displaying the transaction prompt information in the conversation interface. The transaction prompt information may be generated when the comparison result meets the transaction requirements.

[0218] The transaction prompt information may prompt the first user to place an order. For example, the transaction prompt information may be in the form of a pop-up window, which may include an order control. The first user terminal may send an order request to the server terminal in response to the first user's triggering operation on the order control. The transaction prompt information may be different from the second message generated based on the transaction instruction. The second message generated based on the transaction instruction may be only a text message to prompt the first user to complete the transaction, such as "OK, deal".

[0219] For ease of understanding, Figure 4 A schematic diagram of the interface display of a first user terminal in an actual application of an embodiment of the present application is shown. In response to a message sending operation triggered by the first user in a conversation interface 400, the first user terminal displays a first message 41 provided by the first user, "Can this price be discounted? How about 120 yuan?", and sends the first message to the server. The server can recognize that the first user has a bargaining intention based on the first message, and extract the first bid of the first user as 120. Assuming that the price upper limit of the target object is 150 yuan and the lower limit is 110 yuan, and the first bid is within the price range, the server can generate a transaction instruction, and use the second large model to generate a second message 42 such as "Okay, it can be done at 120 yuan" based on the transaction instruction, and send the second message to the first user terminal to be displayed in the conversation interface 400 of the first user terminal. Since the first quotation is within the price range and meets the transaction requirements, the server can send a transaction prompt message 43 to the first user, and the first user can display the transaction prompt message 43 in the conversation interface. The transaction prompt message 43 may include an "order now" control. The first user can respond to the first user's trigger operation on the order control and send an order request to the server. Among them, since the second message is generated by a customer service robot, the word "AI" can be used in the second user's avatar to prompt the first user that he is currently interacting with the customer service robot.

[0220] For ease of understanding, Figure 5 A schematic diagram of scene interaction in an actual application of an embodiment of the present application is shown. Figure 5 As shown, the first user terminal 501 and the second user terminal 502 can respectively establish a connection with the server terminal 503 through the network. The first user terminal 501 and the second user terminal 502 can establish a session channel based on the server, and the first user can interact with messages through the first user terminal 501 and the second user can interact with messages through the second user terminal 502 based on the session channel.

[0221] Among them, the first user terminal 501 and the second user terminal 502 can be a browser, an APP (Application), or a web application such as an H5 (HyperText Markup Language 5, Hypertext Markup Language Version 5) application, or a light application (also known as a mini-program, a lightweight application) or a cloud application, etc. The first user terminal 501 and the second user terminal 502 can be deployed in an electronic device and need to rely on the device to run or some apps in the device to run, etc. The electronic device can, for example, have a display screen and support information browsing, etc., such as a personal mobile terminal such as a mobile phone, a tablet computer, a personal computer, a desktop computer, a smart speaker, a smart watch, etc. For ease of understanding, Figure 5 The first user terminal 501 and the second user terminal 502 are mainly represented by device images. Various other types of applications can usually be configured in electronic devices, such as human-computer dialogue applications, model training applications, text processing applications, web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc. Electronic devices can refer to devices used by users, which have the functions of computing, surfing the Internet, communicating, etc. required by users, such as mobile phones, tablet computers, personal computers, wearable devices, etc. Electronic devices can usually include at least one processing component and at least one storage component. Electronic devices may also include basic configurations such as network card chips, IO buses, audio and video components, which are not limited in this application. Optionally, according to the implementation form of the electronic device, some peripheral devices may also be included, such as keyboards, mice, input pens, printers, etc., which are not limited in this application.

[0222] The server 503 may include servers that provide various services, such as servers that support model training, and servers that process information sent by the user. It should be noted that the server 503 can be implemented as a distributed server cluster consisting of multiple servers, or as a single server. The server can also be a server of a distributed system, or a server combined with a blockchain. The server can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms, or an intelligent cloud computing server or intelligent cloud host with artificial intelligence technology.

[0223] The server 503 may obtain a first message sent by the first user from the first user terminal 501 based on the conversation channel. The server 503 may further process the first message using a customer service robot to generate a second message. The customer service robot may integrate a detection model, an intent recognition model, a first large model, and a second large model.

[0224] Refer to Figure 6 In the bargaining process diagram shown, the server 503 can use the detection model to detect whether the first message is an abnormal message (step 601); if so, the content-friendly predetermined message is used as the generated second message. If not, the recognition operation of the first message is performed, and then the intention recognition model is used to identify whether the first user has a bargaining intention based on the first message (step 602).

[0225] If it is determined that the first user has no bargaining intention, a second message is generated based on the object description information of the target object, the first message, and at least one second historical message. If a reasonable answer can be generated for the first message, a second message containing object information can be generated (step 603). If the user cannot be answered, a content-friendly predetermined message can be used as the generated second message (step 604).

[0226] In the case that the first user has bargaining intention, the first large model can be used to extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel (step 605).

[0227] When the first bid is extracted, if the first bid is higher than the upper price limit of the price range, a rejection instruction is generated, and a second message is generated based on the rejection instruction using the second largest model (step 606); when the first bid does not exceed the price range, a transaction instruction is generated, and a second message is generated based on the transaction instruction using the second largest model (step 607); when the first bid is lower than the lower price limit of the price range, and the upper price limit is different from the lower price limit, a first quotation is selected from the price range and a second message is generated based on the first quotation using the second largest model, or a bid instruction is generated, and a second message is generated based on the bid instruction using the second largest model (step 608); when the first bid is lower than the lower price limit of the price range, and the upper price limit is the same as the lower price limit, a rejection instruction is generated and a second message is generated based on the rejection instruction using the second largest model, or a bid instruction is generated and a second message is generated based on the bid instruction using the second largest model (step 609).

[0228] If the first bid is not extracted, whether a second bid exists is searched from the historical message record of the conversation channel (step 610); if the second bid does not exist, the second bid is selected from the price range, and a second message is generated based on the second bid using the second largest model (step 611);

[0229] In the case where the second bid exists, it is determined whether there is a second bid of the first user in the historical message record. If so, the price range is updated based on the second bid and the second bid, and a third bid is selected from the price range and a second message is generated based on the third bid using the second large model, or a bid instruction is generated and a second message is generated based on the bid instruction using the second large model (step 612); if not, the price range is updated based on the second bid, and a third bid is selected from the price range and a second message is generated based on the third bid using the second large model, or a bid instruction is generated and a second message is generated based on the bid instruction using the second large model (step 613).

[0230] After the server 503 generates the second message, it can send the second message to the first user terminal 501 and the second user terminal 502, and the first user terminal 501 and the second user terminal 502 can display the second message in the corresponding session interface respectively. In addition, if the first bid does not exceed the price range, the server 503 can send a transaction prompt information to the first user terminal 501, and the first user terminal 501 can display the transaction prompt information in the session interface. The first user terminal 501 can generate an order request in response to the first user's trigger operation on the transaction prompt information, and send the order request to the server 503. The server 503 can generate a transaction order based on the order request according to the first bid, and send a success prompt information to the second user terminal 502.

[0231] In this embodiment, the server can generate a second message based on the first message, automatically reply to the message, realize intelligent customer service, ensure communication efficiency and timeliness, and through bargaining intention recognition, bid extraction and price comparison, it can replace the second user to bargain with the first user, avoid the interruption of bargaining communication when the second user is offline, further ensure communication continuity, improve communication efficiency, help improve the transaction rate of the target object, and improve the user's service experience. Among them, the detection model is used to detect whether the first message is an abnormal message, the intention recognition model is used to identify whether the first user has a bargaining intention, the first large model is used to extract the first bid, and the second large model is used to generate the second message. Different types of models are selected for different operations, which can better balance task requirements, resource constraints and performance requirements.

[0232] Figure 7 A schematic diagram of a structure of an embodiment of a service providing device provided in an embodiment of the present application, wherein the method device comprises:

[0233] A first acquisition module, configured to acquire a first message sent by a first user to a target object based on a conversation channel with a second user;

[0234] A first extraction module is used to extract a first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel when identifying that the first user has a bargaining intention based on the first message;

[0235] A first comparison module, configured to compare the first bid with a price range of the target object;

[0236] A generating module, used for generating a second message according to the comparison result;

[0237] The first sending module is used to send the second message to the first user.

[0238] In some embodiments, the generation module may also be used to generate a second message based on the object description information of the target object, the first message, and at least one second historical message when it is identified that the first user has no bargaining intention.

[0239] In some embodiments, the generating module may be specifically configured to generate a deal instruction when the first bid does not exceed the price range, and generate a second message based on the deal instruction;

[0240] generating a rejection instruction if the first bid is higher than the upper price limit of the price range, and generating a second message based on the rejection instruction;

[0241] When the first bid is lower than the lower limit of the price range, and the upper limit of the price is different from the lower limit of the price, selecting the first bid from the price range and generating a second message based on the first bid or generating a bid instruction, and generating a second message based on the bid instruction;

[0242] When the first bid is lower than the lower limit of the price range and the upper limit is the same as the lower limit, a rejection instruction is generated and a second message is generated based on the rejection instruction, or a bid instruction is generated and a second message is generated based on the bid instruction.

[0243] In some embodiments, the device may further include:

[0244] The updating module is used for updating the upper limit of the price range by using the first quotation after the second message is generated, when the first bid is lower than the lower limit of the price range and the first quotation is selected from the price range.

[0245] In some embodiments, after the second message is generated, the updating module may also be configured to select a highest bid from historical bids of the first user to update the lower limit of the price range when the first bid is lower than the lower limit of the price range.

[0246] In some embodiments, the generating module may also be used to search for a second quotation from a historical message record of the session connection without extracting the first bid; and generate a second message based on the search result.

[0247] Optionally, the generating module generates the second message based on the search result, including: if there is no second quotation, selecting the second quotation from the price range, and generating the second message based on the second quotation;

[0248] If there is a second bid, determine whether there is a second bid of the first user in the historical message record.

[0249] If so, updating the price range based on the second quotation and the second bid, and selecting a third quotation from the price range and generating a second message based on the third quotation, or generating a bid instruction and generating a second message based on the bid instruction;

[0250] If not, the price range is updated based on the second quotation, and a third quotation is selected from the price range and a second message is generated based on the third quotation, or a bid instruction is generated and a second message is generated based on the bid instruction.

[0251] In some embodiments, when the first extraction module identifies, based on the first message, that the first user has a bargaining intention, extracting, based on the first message and at least one first historical message corresponding to the session connection, the first bid provided by the first user may include:

[0252] When the customer service robot identifies the first user's intention to bargain based on the first message, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the session connection;

[0253] Generating a second message according to the comparison result includes:

[0254] Based on the comparison result, a second message is generated using the customer service robot.

[0255] In some embodiments, when the first extraction module identifies, based on the first message, that the first user has a bargaining intention, extracting, based on the first message and at least one first historical message corresponding to the session connection, the first bid provided by the first user may include:

[0256] Inputting the first message into an intention recognition model to identify whether the first user has a bargaining intention;

[0257] In the case that the first user has bargaining intention, the first bid provided by the first user is extracted based on the first message and at least one first historical message corresponding to the session connection by using the first large model.

[0258] Optionally, the intention recognition model can be trained in the following manner: the intention recognition model is trained using the second sample data and the training label of the second sample data; the training label of the first sample data includes whether the second sample data has bargaining intention.

[0259] Optionally, the first large model can be obtained by training in the following manner: using the third sample data and the training labels of the third sample data to train the first large model; the training labels of the third sample data include the price data included in the third sample data.

[0260] In some embodiments, when the generation module identifies that the first user has no bargaining intention, generating a second message based on the object description information of the target object, the first message, and at least one second historical message may include: when the generation module identifies that the first user has no bargaining intention, generating a second message based on the object description information of the target object, the first message, and at least one second historical message using a second large model.

[0261] Optionally, generating the second message based on the transaction instruction may include: generating the second message based on the transaction instruction using a second large model;

[0262] Based on the rejection instruction, generating a second message may include: generating a second message based on the rejection instruction using the second large model;

[0263] Generating a second message based on the first quotation may include: generating a second message based on the first quotation using the second large model

[0264] Generating a second message based on the bidding instruction may include: generating a second message based on the bidding instruction using the second large model.

[0265] Optionally, the second largest model can be obtained by training in the following manner: the second largest model is trained using the fourth sample data and the training label of the fourth sample data; the training label of the third sample data includes the response data corresponding to the fourth sample data.

[0266] In some embodiments, the device may further include:

[0267] The detection module is used to detect whether the first message is an abnormal message by using the detection model before the first message is identified. If not, the first extraction module is used to perform the identification operation on the first message; if so, the generation module is used to generate a second message based on the predetermined message.

[0268] In some embodiments, the device may further include:

[0269] The first transaction prompt module is used to send a transaction prompt message for the target object to the first user when the first bid does not exceed the price range; generate a transaction order according to the first bid in response to the first user's order request for the target object; and send a success prompt message to the second user.

[0270] Figure 7 The service provider can execute Figure 1 The implementation principle and technical effect of the service providing method in the embodiment shown are not described in detail. The specific way in which each module and unit performs operations in the service providing device in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0271] Figure 8 A schematic diagram of a structure of an embodiment of a transaction processing device provided in an embodiment of the present application, wherein the method and device comprises:

[0272] A second acquisition module is used to acquire a first message sent by the first user to the target object based on the conversation channel with the second user;

[0273] A second extraction module is configured to extract a first bid of the first user based on the first message and at least one first historical message corresponding to the conversation channel when identifying that the first user has a bargaining intention based on the first message;

[0274] A second comparison module: comparing the first bid with a price range corresponding to the target object;

[0275] The second sending module is configured to send transaction prompt information for the target object to the first user when the comparison result meets the transaction requirements.

[0276] In some embodiments, when the second extraction module identifies that the first user has bargaining intention based on the first message, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel may include: inputting the first message into an intention recognition model to identify whether the first user has bargaining intention; when the first user has bargaining intention, using the first large model to extract the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel.

[0277] Optionally, the intention recognition model can be trained in the following manner: the intention recognition model is trained using the second sample data and the training label of the second sample data; the training label of the first sample data includes whether the second sample data has bargaining intention.

[0278] Optionally, the first large model can be obtained by training in the following manner: the first large model is trained using the third sample data and the training labels of the third sample data; the training labels of the third sample data include the price data included in the third sample data.

[0279] The specific implementation of using the intention recognition model to identify whether the user has bargaining intention and using the first model to extract the first bid can be found in detail. Figure 1 As described in the illustrated embodiment, details will not be repeated here.

[0280] In some embodiments, the device may further include:

[0281] The second transaction prompt module is used to send transaction prompt information for the target object to the first user when the first bid does not exceed the price range; generate a transaction order according to the first bid in response to the first user's order request for the target object; and send a success prompt information to the second user.

[0282] Figure 8 The transaction processing device can execute Figure 2 The implementation principle and technical effect of the transaction processing method described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the transaction processing device in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0283] Fig. 9 A schematic diagram of the structure of an embodiment of an information display device provided in an embodiment of the present application, wherein the method and device comprises:

[0284] A providing module 901 is used to provide a conversation interface between a first user and a second user.

[0285] The response module 902 is used to determine a first message provided by the first user and directed to a target object in response to a message sending operation triggered by the first user in the conversation interface.

[0286] The third sending module 903 is used to send a first message to the server, so that the server can identify, based on the first message, that the first user has the intention to bargain, and based on the first message and at least one first historical message corresponding to the conversation channel between the first user and the second user, extract the first bid provided by the first user; compare the first bid with the price range of the target object; and generate a second message according to the comparison result.

[0287] Display module 904: obtains the second message, and displays the second message in the conversation interface.

[0288] In some embodiments, the display module may also be used to: obtain transaction prompt information sent by the server; and display the transaction prompt information in the session interface.

[0289] Fig. 9 The information display device can perform Figure 3 The implementation principle and technical effect of the information display method described in the illustrated embodiment will not be described in detail. The specific manner in which each module and unit performs operations in the information display device in the above embodiment has been described in detail in the embodiment of the method, and will not be described in detail here.

[0290] The present application also provides a computing device, such as Fig.10 As shown, the device may include a storage component 1001 and a processing component 1002;

[0291] The storage component 1001 stores one or more computer instructions, wherein the one or more computer instructions are called and executed by the processing component to implement the following Figure 1 The service providing method described in the embodiment shown or Figure 2 The transaction processing method described in the illustrated embodiment.

[0292] Of course, the computing device may also include other components, such as input / output interfaces, display components, communication components, etc.

[0293] The display component in the computing device is used to implement the following Figure 3 In the case of the information display method described in the illustrated embodiment, the computing device may further include a display component to perform corresponding display operations.

[0294] The input / output interface provides an interface between the processing component and the peripheral interface module, which may be an output device, an input device, etc. The communication component is configured to facilitate wired or wireless communication between the computing device and other devices.

[0295] The processing component 1002 may include one or more processors to execute computer instructions to complete all or part of the steps in the above method. Of course, the processing component may also be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic components to perform the above method.

[0296] The storage component 1001 is configured to store various types of data to support operations at the terminal. The storage component can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk.

[0297] The display component may be an electroluminescent (EL) element, a liquid crystal display or a micro display having a similar structure, or a retinal direct display or a similar laser scanning display.

[0298] It should be noted that the above computing device can be a physical device or an elastic computing host provided by a cloud computing platform, etc. 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.

[0299] It should be noted that the above computing device is implemented as follows Figure 1 The service providing method described in the embodiment shown or Figure 2 In the case of the transaction processing method described in the illustrated embodiment, it can be a physical device or an elastic computing host provided by a cloud computing platform. 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. The above computing device implements Figure 3 In the case of the information display method described in the illustrated embodiment, it can be specifically implemented as an electronic device. The electronic device can refer to a device used by a user and having the computing, Internet access, communication and other functions required by the user, such as a mobile phone, tablet computer, personal computer, wearable device, etc.

[0300] The present application also provides a computer-readable storage medium storing a computer program, wherein the computer program can achieve the above-mentioned Figure 3 The information display method described in the embodiment shown or Figure 1 The service providing method described in the embodiment shown or Figure 2 The transaction processing method described in the illustrated embodiment. The computer-readable medium may be included in the electronic device described in the above embodiment; or it may exist independently without being assembled into the electronic device.

[0301] The present application also provides a computer program product, which includes a computer program carried on a computer-readable storage medium. When the computer program is executed by a computer, the computer program can achieve the above-mentioned Figure 3 The information display method described in the embodiment shown or Figure 1The service providing method described in the embodiment shown or Figure 2 The transaction processing method described in the illustrated embodiment. In such an embodiment, the computer program can be downloaded and installed from a network, and / or installed from a removable medium. When the computer program is executed by a processor, various functions defined in the system of the present application are executed.

[0302] In the above corresponding embodiments, the computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device, or device.

[0303] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0304] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0305] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0306] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A service providing method, characterized in that: include: Acquire a first message sent by the first user to the target object based on the conversation channel with the second user; Based on the first message, if it is identified that the first user has a bargaining intention, extracting a first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel; comparing the first bid to a price range of the target object; generating a second message according to the comparison result; The second message is sent to the first user.

2. The method according to claim 1, characterized in that Also includes: When it is identified that the first user has no bargaining intention, a second message is generated based on the object description information of the target object, the first message and at least one second historical message.

3. The method according to claim 1, characterized in that Generating a second message according to the comparison result includes: If the first bid does not exceed the price range, generating a transaction instruction, and generating a second message based on the transaction instruction; If the first bid is higher than the upper limit of the price range, generating a rejection instruction, and generating a second message based on the rejection instruction; When the first bid is lower than the lower limit of the price range, and the upper limit of the price is different from the lower limit of the price, selecting the first bid from the price range and generating a second message based on the first bid or generating a bid instruction, and generating a second message based on the bid instruction; When the first bid is lower than the lower limit of the price range and the upper limit is the same as the lower limit, a rejection instruction is generated and a second message is generated based on the rejection instruction, or a bid instruction is generated and a second message is generated based on the bid instruction.

4. The method according to claim 3, characterized in that After the second message is generated, the method further includes: In a case where the first bid is lower than a lower price limit of the price range and a first offer is selected from the price range, an upper price limit of the price range is updated using the first offer.

5. The method according to claim 3 or 4, characterized in that: After the second message is generated, the method further includes: In the case where the first bid is lower than the lower limit of the price range, the highest bid is selected from the historical bids of the first user to update the lower limit of the price range.

6. The method according to claim 1, characterized in that Also includes: In the case where the first bid is not obtained, searching from the historical message record of the conversation channel whether there is a second bid; Based on the search result, a second message is generated.

7. The method according to claim 6, characterized in that The generating a second message based on the search result includes: If no second quotation exists, selecting a second quotation from the price range, and generating a second message based on the second quotation; In the case where there is a second bid, determining whether there is a second bid of the first user in the historical message record, If so, updating the price range based on the second quotation and the second bid, and selecting a third quotation from the price range and generating a second message based on the third quotation, or generating a bid instruction and generating a second message based on the bid instruction; If not, the price range is updated based on the second quotation, and a third quotation is selected from the price range and a second message is generated based on the third quotation, or a bid instruction is generated and a second message is generated based on the bid instruction.

8. The method according to claim 1, characterized in that In the case where, based on the first message, it is identified that the first user has a bargaining intention, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel includes: When the customer service robot identifies the first user's intention to bargain based on the first message, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel; Generating a second message according to the comparison result includes: Based on the comparison result, a second message is generated using the customer service robot.

9. The method according to claim 1, characterized in that: In the case where, based on the first message, it is identified that the first user has a bargaining intention, extracting the first bid provided by the first user based on the first message and at least one first historical message corresponding to the conversation channel includes: Inputting the first message into an intention recognition model to identify whether the first user has a bargaining intention; In the case that the first user has bargaining intention, a first bid provided by the first user is extracted based on the first message and at least one first historical message corresponding to the conversation channel by using a first large model.

10. The method according to claim 9, characterized in that The intent recognition model is trained in the following way: The intention recognition model is trained using the second sample data and the training label of the second sample data; the training label of the second sample data includes whether the second sample data has bargaining intention.

11. The method according to claim 9, characterized in that The first model is trained in the following way: The first large model is trained using the third sample data and the training labels of the third sample data; the training labels of the third sample data include the price data included in the third sample data.

12. The method according to claim 2, characterized in that: The generating the second message based on the object description information of the target object, the first message and at least one second historical message when it is identified that the first user has no bargaining intention includes: When it is identified that the first user has no bargaining intention, a second message is generated using a second large model based on the object description information of the target object, the first message and at least one second historical message.

13. The method according to claim 3, characterized in that The generating a second message based on the transaction instruction includes: generating a second message based on the transaction instruction using a second large model; The generating a second message based on the rejection instruction includes: generating a second message based on the rejection instruction using the second large model; The generating a second message based on the first quotation includes: generating a second message based on the first quotation using the second large model The generating a second message based on the bidding instruction includes: generating a second message based on the bidding instruction using the second large model.

14. The method according to claim 13, characterized in that The second largest model is trained as follows: The second large model is trained using the fourth sample data and the training labels of the fourth sample data; the training labels of the fourth sample data include the response data corresponding to the fourth sample data.

15. The method according to claim 1, characterized in that Before the first message is identified, the method further includes: Using the detection model, detecting whether the first message is an abnormal message; If not, performing an identification operation on the first message; If so, a second message is generated based on the predetermined message.

16. The method according to claim 3, characterized in that Also includes: When the first bid does not exceed the price range, sending transaction prompt information for the target object to the first user; In response to the first user's order request for the target object, generating a transaction order according to the first bid; A success prompt message is sent to the second user.

17. A transaction processing method, characterized in that: include: Acquire a first message sent by the first user to the target object based on the conversation channel with the second user; Based on the first message, if it is identified that the first user has a bargaining intention, extracting a first bid of the first user based on the first message and at least one first historical message corresponding to the conversation channel; comparing the first bid with a price range corresponding to the target object; When the comparison result meets the transaction requirements, a transaction reminder message for the target object is sent to the first user.

18. An information display method, characterized in that: include: Providing a conversation interface between the first user and the second user; In response to a message sending operation triggered by the first user on the conversation interface, determining a first message provided by the first user and directed to a target object; sending the first message to a server, so that the server can identify, based on the first message, that the first user has a bargaining intention, and extract, based on the first message and at least one first historical message corresponding to a conversation channel between the first user and the second user, a first bid provided by the first user; comparing the first bid with the price range of the target object; and generating a second message according to the comparison result; The second message is obtained, and the second message is displayed in the conversation interface.

19. The method according to claim 18, characterized in that Also includes: Acquiring transaction prompt information sent by the server; the transaction prompt information is generated when the comparison result meets the transaction requirements; The transaction prompt information is displayed in the conversation interface.

20. A computing device, characterized in that It comprises a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are used to be called and executed by the processing component to implement the service providing method as described in any one of claims 1 to 16 or the transaction processing method as described in claim 17 or the information display method as described in any one of claims 18 to 19.

21. A computer storage medium, characterized in that A computer program is stored, and when the computer program is executed by a computer, the service providing method according to any one of claims 1 to 16, the transaction processing method according to claim 17, or the information display method according to any one of claims 18 to 19 is implemented.

22. A computer program product, characterized in that It includes a computer program / instruction, which, when executed by a computer, implements the service providing method according to any one of claims 1 to 16, the transaction processing method according to claim 17, or the information display method according to any one of claims 18 to 19.

Images