Information processing method, device and equipment

Abstract

The embodiment of the invention discloses an information processing method, device and equipment, and the method comprises the steps: receiving complaint information submitted by a first user, and the complaint information comprises information interaction record information between the first user and a second user; converting the information interaction record information into information interaction sequence data which can be input into an intention recognition model; inputting the information interaction sequence data into a coding module in an intention recognition model to obtain coding information corresponding to the information interaction sequence data, the coded information corresponding to the information interaction sequence data at least comprises sub-coded information of a user to which each character in the information interaction sequence data belongs and sub-coded information corresponding to identification information of the second user set for interaction information of each second user; and inputting the coding information into an intention recognition module in an intention recognition model to obtain intention information corresponding to the information interaction record information, and further determining a complaint response strategy for complaint information submitted by the first user.

Description

Technical Field

[0001] This document relates to the field of computer technology, and in particular to an information processing method, device and equipment. Background Art

[0002] In many business scenarios (such as intelligent adjudication business, adjudication business for privacy data protection, risk prevention and control business, etc.), when a user encounters a certain risk, he or she will file a complaint with the corresponding business provider. For example, in the intelligent adjudication business scenario, after a user is defrauded or faces the risk of fraud, he or she will file a complaint request at the reporting center set up by the provider of the intelligent adjudication business, and submit complaint information at the same time, which will include a description of how the user was defrauded and details, as well as complaint voucher information, etc. Among them, one of the types of complaint voucher information submitted by the user is chat records or conversation records, such as chat records of instant messaging applications, conversation records of financial transaction applications, etc. The above chat records or conversation records often contain a lot of information interaction content and a large amount of information. When conducting a qualitative analysis of the user's complaint, the process of the user being defrauded or the risk of fraud can be restored based on the above chat records or conversation records, and the risk can be qualitatively analyzed at the same time. However, the information interaction in chat records or conversation records is often formed by multiple rounds of dialogue, and is different from the real-time interaction, in which the two parties take turns to ask and answer questions in a multi-round dialogue format. Instead, it is a form without a fixed dialogue order. Therefore, it is necessary to provide a better multi-round information interaction intention recognition mechanism to identify the dialogue subjects in the chat records or conversation records, restore the information interaction logic, and then perform intent recognition processing. Summary of the invention

[0003] The purpose of this specification is to provide a better multi-round information interaction intention recognition mechanism, so as to identify the conversation subjects in the chat records or conversation records, restore the information interaction logic, and then perform intent recognition processing.

[0004] In order to realize the above technical solution, this specification is implemented as follows: This specification provides an information processing method, the method comprising: receiving complaint information submitted by a first user, the complaint information including information interaction record information between the first user and a second user. Converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model. Inputting the information interaction sequence data into an encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module, and obtaining encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-encoding information of the user to which each character in the information interaction sequence data belongs and sub-encoding information corresponding to the identification information of the second user set for each interaction information of the second user. Inputting the encoding information corresponding to the information interaction sequence data into the intention recognition module in the intention recognition model, obtaining the intention information corresponding to the information interaction record information, and determining a complaint response strategy for the complaint information submitted by the first user based on the intention information corresponding to the information interaction record information to respond to the complaint information of the first user.

[0005] This specification provides an information processing device, the device comprising: an information receiving module, receiving complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user. A conversion module, converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model. An encoding module, inputting the information interaction sequence data into an encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module, and obtaining encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-encoding information of the user to which each character in the information interaction sequence data belongs and sub-encoding information corresponding to the identification information of the second user set for each interaction information of the second user. An intention determination module, inputting the encoding information corresponding to the information interaction sequence data into an intention recognition module in the intention recognition model, obtaining the intention information corresponding to the information interaction record information, and determining a complaint response strategy for the complaint information submitted by the first user based on the intention information corresponding to the information interaction record information to respond to the complaint information of the first user.

[0006] The present specification provides an information processing device, the information processing device comprising: a processor; and a memory arranged to store computer executable instructions, the executable instructions, when executed, enable the processor to: receive complaint information submitted by a first user, the complaint information including information interaction record information between the first user and a second user. Convert the information interaction record information into information interaction sequence data that can be input into an intention recognition model. Input the information interaction sequence data into an encoding module in the intention recognition model, and encode the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-encoding information of the user to which each character in the information interaction sequence data belongs and sub-encoding information corresponding to the identification information of the second user set for each interaction information of the second user. Input the encoding information corresponding to the information interaction sequence data into the intention recognition module in the intention recognition model, obtain the intention information corresponding to the information interaction record information, and determine the complaint response strategy for the complaint information submitted by the first user based on the intention information corresponding to the information interaction record information to respond to the complaint information of the first user.

[0007] The present specification also provides a storage medium, the storage medium is used to store computer executable instructions, and the executable instructions implement the following process when executed by the processor: receiving complaint information submitted by a first user, the complaint information including information interaction record information between the first user and the second user. Converting the information interaction record information into information interaction sequence data that can be input into the intention recognition model. Inputting the information interaction sequence data into the encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module, and obtaining the encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user. Inputting the encoding information corresponding to the information interaction sequence data into the intention recognition module in the intention recognition model, obtaining the intention information corresponding to the information interaction record information, and determining the complaint response strategy for the complaint information submitted by the first user to respond to the complaint information of the first user based on the intention information corresponding to the information interaction record information.

[0008] The present specification also provides a computer program product, including a computer program, which implements the following process when executed by a processor: receiving complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user. Converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model. Inputting the information interaction sequence data into an encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module, and obtaining encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-encoding information of the user to which each character in the information interaction sequence data belongs and sub-encoding information corresponding to the identification information of the second user set for each interaction information of the second user. Inputting the encoding information corresponding to the information interaction sequence data into the intention recognition module in the intention recognition model, obtaining the intention information corresponding to the information interaction record information, and determining a complaint response strategy for the complaint information submitted by the first user to respond to the complaint information of the first user based on the intention information corresponding to the information interaction record information. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of this specification or the prior art, the drawings required for use in the embodiments or the prior art description will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative labor. Figure 1 This is an information processing method embodiment of this specification; Figure 2 This is a schematic diagram of an information interaction record in this manual; Figure 3 This is a schematic diagram of a coding process for this specification; Figure 4 This is a schematic diagram of another encoding process of this specification; Figure 5 This is another information processing method embodiment of the present specification; Figure 6 This is another information processing method embodiment of the present specification; Figure 7 This is an information processing device embodiment of the present specification; Figure 8 This is an information processing device embodiment of this specification. DETAILED DESCRIPTION

[0010] The embodiments of this specification provide an information processing method, apparatus and device.

[0011] In order to enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below in conjunction with the drawings in the embodiments of this specification. Obviously, the described embodiments are only part of the embodiments of this specification, not all of the embodiments. Based on the embodiments in this specification, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this specification.

[0012] The embodiments of this specification provide a multi-round information interaction intention recognition mechanism based on subject information. In many business scenarios (such as intelligent trial business, risk prevention and control business, etc.), when a user encounters a certain risk, he will file a complaint with the corresponding business provider. For example, in the intelligent trial business scenario, after the user is defrauded or the risk of fraud occurs, he will file a complaint request in the reporting center set up by the provider of the intelligent trial business, and submit complaint information at the same time, which will include a description of how the fraud occurred and details, as well as complaint voucher information, etc. Among them, one of the types of complaint voucher information submitted by the user is a chat record or a conversation record, such as a chat record of an instant messaging application, a conversation record of a financial transaction application, etc. The above chat records or conversation records often contain a lot of information interaction content and a large amount of information. When conducting a qualitative analysis of the user's complaint, the process of the user being defrauded or the risk of fraud can be restored based on the above chat records or conversation records, and the risk can be qualitatively analyzed at the same time. However, the information interaction in the chat record or dialogue record is often formed by multiple rounds of dialogue, and it is different from the multi-round dialogue form of real-time interaction, in which the two subjects take turns to ask and answer questions, but it is a form without a fixed dialogue order. Therefore, it is necessary to first identify the dialogue subject in the chat record or dialogue record, restore the information interaction logic, and then perform the intention recognition process. The embodiment of this specification provides a multi-round information interaction intention recognition mechanism based on subject information, so as to identify the dialogue subject in the chat record or dialogue record, restore the information interaction logic, and then perform the intention recognition process. For specific processing, please refer to the specific content in the following embodiment.

[0013] like Figure 1As shown, an embodiment of this specification provides an information processing method, and the execution subject of the method can be a terminal device or a server, etc., wherein the terminal device can be a mobile terminal device such as a mobile phone, a tablet computer, or a computer device such as a laptop or a desktop computer, or an IoT device (specifically such as a smart watch, a car device, etc.), etc., wherein the server can be an independent server, or a server cluster composed of multiple servers, etc., and the server can be a background server for financial services or online shopping services, or a background server for an application, etc. In this embodiment, the execution subject is a server as an example for detailed description. For the case where the execution subject is a terminal device, please refer to the following server situation processing, which will not be repeated here. The method can specifically include the following steps: In step S102, complaint information submitted by a first user is received, where the complaint information includes information interaction record information between the first user and a second user.

[0014] Among them, the complaint information can be the relevant information of a complaint against a certain transaction, wherein the transaction can be any transaction, such as payment, transfer, etc. in financial services, or insurance transactions or other forms of equivalent exchange, etc. In addition, the complaint information can also be the relevant information of a complaint against a certain penalty against a user or a restriction on the use of a certain function. At this time, the complaint information can be used to remove restrictions on the user's account, etc., which can be set according to actual conditions, and the embodiments of this specification do not limit this. The first user can be any user who initiates a complaint request. For a complaint against a certain transaction, the first user can be one of the two parties to the transaction, that is, the purchaser, or the merchant, etc. For a complaint used to remove restrictions on a certain account, the first user can be the party that is punished or the party whose use of a certain function is restricted. The second user can be the other party involved in the complaint information, such as the fraudulent party or the user who applies for a certain penalty against the first user or a restriction on the use of a certain function, etc., which can be set according to actual conditions. The information interaction record information may be a chat record, a conversation record, etc. The information interaction record information may be carried in a variety of different ways. For example, the above-mentioned information interaction record information may be carried through text, or through images, such as screenshots of chat records, etc. The above-mentioned information interaction record information may also be carried through audio, such as voice of a conversation, etc. The above-mentioned information interaction record information may also be carried through video, such as a video of a conversation, etc. It may be set specifically according to actual conditions, and the embodiments of this specification do not limit this.

[0015] In implementation, the complaint information submitted by the first user can be received for different application scenarios. For example, for a complaint about a transaction, the buyer can select the goods to be purchased in the merchant's store (such as physical goods, virtual goods or services, etc.). After selecting the goods, the buyer scans the code or triggers the specified financial service (such as payment or transfer, etc.) in the application provided by the merchant. The buyer's terminal device calls up the corresponding interface of the financial application and completes the processing of the financial service to obtain the goods. The merchant can distribute the goods through the designated network channel or by mail according to the goods purchased by the buyer. In the process of the above transaction, the buyer and the merchant can communicate about the goods or the transaction, and the transaction can be completed after the two parties reach an agreement. If one party to the transaction finds that there may be a certain risk, it can initiate a complaint on the transaction platform (i.e., server) of the above transaction, and a complaint mechanism can be set up in the transaction platform. Specifically, a "complaint center" (or "report center") or a complaint entrance (such as a complaint link or complaint button, etc.) can be set up in the transaction platform. The first user can select the transaction to be complained in the "complaint center". At the same time, the first user can enter the content to be complained, the complaint risk type of the transaction, the entire event, etc., and can upload the complaint voucher information, such as the relevant text description, image, audio data or video data of the information interaction record of the two parties in the transaction. In addition, the complaint voucher information can also include screenshots of relevant pages (such as transfer pages, product delivery status pages, etc.) during the transaction, etc., which can be set according to actual conditions. After the above information is entered, the "complaint" button can be clicked. At this time, the terminal device can obtain the above information selected or entered by the user, and can generate the complaint information of the first user based on the obtained information, and can further generate a complaint request based on the complaint information, and can submit the complaint request to the server, so that the server can receive the complaint information submitted by the first user.

[0016] For complaints for lifting restrictions on a certain account, if the first user finds that he has been punished or the use of a certain function has been restricted, he can determine the events that may be controversial or problematic (such as transfer events or transaction events, etc.) that existed before the time point of being punished or the use of a certain function being restricted, and can collect data related to the event, which may include information exchange records between the first user and the second user. Then, a complaint or lifting request can be initiated in the set "complaint center" (or "restriction removal center"), and the first user can select the transaction to be complained in the "complaint center". At the same time, the content to be complained, the entire event process, etc., can be entered, and the complaint voucher information can be uploaded, such as the relevant text description, image, audio data or video data of the information exchange record of the two parties in the transaction, which can be set according to the actual situation. After the above information is entered, the "complaint" button can be clicked. At this time, the terminal device can obtain the above information selected or entered by the user, and can generate the complaint information of the first user based on the obtained information, and can further generate a complaint request based on the complaint information, and can submit the complaint request to the server, so that the server can receive the complaint information submitted by the first user.

[0017] It should be noted that the above-mentioned complaint information received from the first user includes information on information interaction records between the first user and the second user. When the information interaction records between the first user and the second user are located in images, audio data or video data, the content of the information interaction can be extracted from the image, audio data or video data through a designated information extraction component (such as an OCR component, etc.) or a model (such as a neural network model or a large language model, etc.), and the content can be converted into text data, thereby obtaining the information interaction record information between the first user and the second user. The specific setting can be based on actual conditions.

[0018] In step S104, the above information interaction record information is converted into information interaction sequence data that can be input into the intention recognition model.

[0019] Among them, the intent recognition model can be used to determine the actual intention of the information interaction record information submitted by the user, for example, through the information interaction record information submitted by the user, the intent recognition model is used to determine whether the user is defrauded, or, through the information interaction record information submitted by the user, the intent recognition model is used to determine whether the user's account is punished or the use of a certain function is restricted. Whether it is reasonable (or whether it complies with corresponding rules, etc.). The intent recognition model can be constructed by a variety of different algorithms or networks. For example, the intent recognition model can be constructed by feature extraction algorithms, clustering algorithms, similarity algorithms, etc., or the intent recognition model can also be constructed by convolutional neural networks, recurrent neural networks, etc., or the intent recognition model can also be constructed by a specified algorithm and network. Specifically, the intent recognition model can be constructed by feature extraction algorithms, convolutional neural networks, etc., which can be set according to actual conditions, and the embodiments of this specification do not limit this.

[0020] In implementation, the information interaction record information obtained above can be analyzed according to the requirements of the input data of the intention recognition model. In this embodiment, the character sequence of each interactive information in the information interaction record information can be converted into a tag sequence according to the requirements of the input data of the intention recognition model. For example, Figure 2As shown, the information interaction record includes 9 pieces of interaction information, namely, "Can I activate it tomorrow? I don't want to wait for a month like others", "No", "Be patient", "OK", "Just saw someone say he has been waiting for a month", "It will be activated for you after the connection", "Please wait patiently", "Sorry, I misread it before", "It's okay, don't worry, it won't happen", and the above 9 pieces of interaction information can be spliced ​​to obtain a character sequence of "Can I activate it tomorrow? I don't want to wait for a month like others. No. Be patient. OK. Just saw someone say he has been waiting for a month. It will be activated for you after the connection. Please wait patiently. Sorry, I misread it before. It's okay, don't worry, it won't happen." The above character sequence can be converted into a tag sequence, that is, [CLS] Can I activate it tomorrow? I don't want to wait for a month like others [SEP] No [SEP] Please be patient [SEP] OK [SEP] Mainly, I just saw someone say that he waited for a month [SEP] Activate it for you after docking [SEP] Please wait patiently [SEP] Sorry, I misread it before [SEP] It's okay, don't worry. No, the above tag sequence includes the sub-tag sequence corresponding to the character sequence of each interactive information, namely [CLS] Can I activate it tomorrow? I don't want to wait for a month like others, [SEP] No, [SEP] Please be patient, [SEP] OK, [SEP] Mainly, I just saw someone say that he waited for a month, [SEP] Activate it for you after docking, [SEP] Please wait patiently, [SEP] Sorry, I misread it before, [SEP] It's okay, don't worry. No.

[0021] In step S106, the information interaction sequence data is input into the encoding module in the intention recognition model, and the information interaction sequence data is encoded by the encoding module to obtain encoding information corresponding to the information interaction sequence data. The encoding information corresponding to the information interaction sequence data includes at least sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the second user's identification information set for the interaction information of each second user.

[0022] Among them, the encoding module can be constructed by a variety of different algorithms or networks. For example, the encoding module can be constructed by a neural network or a Transformer network, or it can also be constructed by a mapping algorithm, etc., which can be set according to actual conditions. Through the encoding module, the characters can be converted into corresponding matrices or vectors (such as embedding vectors, etc.), etc., which can be set according to actual conditions. The identification information of the second user can include multiple types. For example, the identification information of the second user can be the nickname or user name information set by the second user in the instant messaging application, or the nickname or user name information set by the server for the second user, or the name information of other users for the second user. In this case, the second user can include two information representing its identity, namely the nickname (or user name) information of the second user and the name information of other users for the second user. In actual applications, the above two information representing its identity can be used as the identification information of the second user, which can be set according to actual conditions.

[0023] In implementation, the information interaction sequence data is input into the encoding module in the intent recognition model, and the information interaction sequence data is encoded by the encoding module. In the process of encoding the information interaction sequence data by the encoding module, a variety of different encoding mechanisms can be set. First, the characters in the information interaction sequence data can be encoded. Specifically, the mapping relationship between different characters and Embedding vectors can be pre-set. For a certain character in the information interaction sequence data, the Embedding vector corresponding to the character can be determined through the above mapping relationship. In the above manner, the Embedding vector corresponding to each character can be obtained, thereby completing the encoding processing of the characters in the information interaction sequence data. Alternatively, a coding algorithm can be pre-set, and the coding information of each character, word or sentence can be calculated through the coding algorithm to complete the encoding processing of the characters in the information interaction sequence data.

[0024] Secondly, since the information interaction process often involves multiple different users, and a piece of interactive information often corresponds to only one user, the same piece of interactive information, if belonging to different users, often has different meanings. Therefore, the user to which each character in the information interaction sequence data belongs can be encoded. For example, Figure 2As shown, the first user: Can I activate tomorrow? I don't want to wait for a month like others; the second user: No; the second user: Be patient, then each character in "Can I activate tomorrow? I don't want to wait for a month like others" belongs to the first user, then the first user can be set as user 1, for each of the above characters, "user 1" is encoded respectively to obtain corresponding sub-coding information, at this time, each of the above characters corresponds to the encoding information of "user 1"; each character in "No" and "Be patient" belongs to the second user, then the second user can be set as user 2, for each character in "No" and "Be patient", "user 2" is encoded respectively to obtain corresponding sub-coding information, at this time, each character in "No" and "Be patient" corresponds to the encoding information of "user 2". In addition, the user's identification information can often reflect the association relationship between different users. Therefore, the identification information of the user to which each interactive information in the information interaction sequence data belongs can be encoded. For example, based on the above example, the two interactive information "No" and "Be patient" both belong to the second user. At this time, for the above two interactive information, the identification information of the second user can be encoded to obtain the encoded information of the identification information of the second user. At this time, the above two interactive information both correspond to the encoded information of the identification information of the second user. In addition, "Can I activate it tomorrow? I don't want to wait for a month like others" belongs to the first user. For the above interactive information "Can I activate it tomorrow? I don't want to wait for a month like others", the identification information of the first user can also be obtained, and the identification information of the first user can be encoded to obtain the first user. The coding information of the user's identification information, at this time, the above-mentioned interactive information "Can I activate it tomorrow? I don't want to wait for a month like others" corresponds to the coding information of the first user's identification information, wherein the interactive information corresponding to the coding information of the user's identification information can be set at the starting position of the interactive information, or can be set at the end position of the interactive information, and can also be set to correspond to each character in the interactive information, that is, the first character in the interactive information corresponds to the coding information of the user's identification information, the second character in the interactive information also corresponds to the coding information of the user's identification information, the third character in the interactive information also corresponds to the coding information of the user's identification information, ..., the last character in the interactive information also corresponds to the coding information of the user's identification information, etc., which can be set specifically according to actual conditions, and the embodiments of this specification do not limit this.

[0025] Afterwards, since characters at different positions or in different contexts may express different meanings, the positions of the characters may be encoded to obtain sub-coding information of the position of each character in the information interaction sequence data.

[0026] Finally, considering that the meanings of sentences expressed in different contexts are often different, for this purpose, a corresponding sentence identifier or paragraph identifier can be set for each sentence, and the sentence identifier or paragraph identifier corresponding to each sentence in the information interaction sequence data can be encoded to obtain the encoding information of the sentence identifier or paragraph identifier corresponding to each sentence. The encoding information of the sentence identifier or paragraph identifier corresponding to the sentence can be set at the starting position of the interactive information where the sentence is located, or at the end position of the interactive information, or can be set to correspond to each character in the sentence, that is, the first character in the sentence corresponds to the encoding information of the sentence identifier or paragraph identifier corresponding to the sentence, the second character in the sentence also corresponds to the encoding information of the sentence identifier or paragraph identifier corresponding to the sentence, the third character in the sentence also corresponds to the encoding information of the sentence identifier or paragraph identifier corresponding to the sentence, ..., the last character in the sentence also corresponds to the encoding information of the sentence identifier or paragraph identifier corresponding to the sentence, etc. It can be set specifically according to actual conditions, and the embodiments of this specification do not limit this.

[0027] It should be noted that the above-mentioned specific encoding process is only an optional processing method, and the information interaction sequence data can also be encoded in other ways, which can be set according to actual conditions. In addition, the above-mentioned division of the encoding process of the information interaction sequence data into the encoding process of characters, the encoding process of the positions of characters, and the encoding process of sentences or paragraphs is only an optional division method and a corresponding encoding process. In actual applications, the encoding process of the information interaction sequence data can also be divided by other division methods. For example, in addition to encoding the information of the user to which each character in the information interaction sequence data belongs and encoding the identification information of the second user set for the interactive information of each second user, it is only necessary to encode the characters, or only encode the information of the user to which each character in the information interaction sequence data belongs and encode the identification information of the second user set for the interactive information of each second user. The encoding process includes encoding the identification information of the second user set for the interactive information of each second user, or, in addition to encoding the information of the user to which each character in the information interaction sequence data belongs and encoding the identification information of the second user set for the interactive information of each second user, only the encoding process of the character and the encoding process of the position of the character are required, in addition to encoding the information of the user to which each character in the information interaction sequence data belongs and encoding the identification information of the second user set for the interactive information of each second user, only the encoding process of the position of the character is required, etc., which can be specifically set according to actual conditions, and the embodiments of this specification are not limited to this.

[0028] Through the above-mentioned various encoding processes, corresponding sub-coding information (i.e., sub-coding information of encoding processing for characters, sub-coding information of encoding processing for character positions, sub-coding information of encoding processing for sentences or paragraphs, sub-coding information of users to which each character in the information interaction sequence data belongs, and sub-coding information corresponding to identification information of the second user set for the interactive information of each second user) can be obtained respectively. The above-mentioned sub-coding information can be fused (such as adding sub-coding information or weighted summing sub-coding information or splicing sub-coding information, etc.), and finally, the encoding information corresponding to the information interaction sequence data can be obtained.

[0029] In step S108, the encoded information corresponding to the information interaction sequence data is input into the intention recognition module in the intention recognition model to obtain the intention information corresponding to the information interaction record information, and based on the intention information corresponding to the information interaction record information, a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user.

[0030] Among them, the intent recognition module can be constructed through a variety of different algorithms or networks. For example, the intent recognition module can be constructed through a neural network or a Transformer network, or it can also be constructed through a specified algorithm (such as a clustering algorithm, a classification algorithm, etc.), etc., which can be set according to the actual situation. Through the encoding module, based on the encoding information corresponding to the information interaction sequence data, the corresponding intent information can be determined, which can be set according to the actual situation. The intent information can include a variety of information, such as whether there is fraud, whether to carry out anti-illegal financial activities, etc., which can be set according to the actual situation.

[0031] The architecture of the intention recognition module in the intention recognition model can be constructed by the above algorithm or network, and then the corresponding training samples can be obtained, and the intention recognition module can be trained using the training samples to obtain the trained intention recognition module. The coded information corresponding to the information interaction sequence data can be input into the intention recognition module in the intention recognition model, and the coded information corresponding to the information interaction sequence data can be processed by the intention recognition module to obtain the corresponding result, which is the intention information corresponding to the information interaction record information. Subsequently, the intention information corresponding to the information interaction record information can be used to determine the complaint response strategy for the complaint information submitted by the first user. For example, if the intention information corresponding to the information interaction record information is fraud, the second user can be subjected to risk prevention and control, such as freezing the second user's account, etc. If the intention information corresponding to the information interaction record information is to lift restrictions on anti-illegal financial activities, and the first user's account can be lifted, the first user's account can be lifted, etc., and the complaint information of the first user can be responded to based on the above complaint response strategy.

[0032] The embodiment of the present specification provides an information processing method, which receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user. Then, the information interaction record information can be converted into information interaction sequence data that can be input into an intention recognition model. After that, the information interaction sequence data can be input into an encoding module in the intention recognition model. The information interaction sequence data is encoded by the encoding module to obtain encoding information corresponding to the information interaction sequence data. The encoding information corresponding to the information interaction sequence data includes at least sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user. Finally, the encoding information corresponding to the information interaction sequence data can be input into the intention recognition module in the intention recognition model to obtain intent information corresponding to the information interaction record information. The intention information corresponding to the mutually recorded information is determined, and a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. In this way, the identification information of the user in the information interaction process (such as the user's nickname information and / or remark identification information, etc.) is introduced, and the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user are set in the coding information corresponding to the information interaction sequence data, so as to distinguish the subjects in the information interaction (the person himself or the active party, the other party or the passive party), and then restore the information interaction logic, and then perform intent recognition processing, which helps the intent recognition model to better understand the expression content in the information interaction. For the back and forth of information in the information interaction, the scene of repeated mention of keywords can effectively understand and recognize its true intention, improve the understanding of the identities of both parties in the information interaction, and improve the accuracy of intent recognition in the information interaction.

[0033] In practical applications, the above-mentioned encoding module includes an encoding module constructed by a Transformer network (or Transformer Block), and accordingly, the encoding information corresponding to the information interaction sequence data also includes information interaction sequence sub-coding information corresponding to the information interaction sequence data and sub-coding information of the position corresponding to each character in the information interaction sequence data.

[0034] Among them, the information interaction sequence data can be called a token sequence, each character in the information interaction sequence data is a token, and the information interaction sequence sub-coding information corresponding to the information interaction sequence data can be represented by an Embedding vector, that is, the information interaction sequence sub-coding information corresponding to the information interaction sequence data can be TokenEmbeddings (i.e., a token embedding vector). Correspondingly, the sub-coding information of the position corresponding to each character in the information interaction sequence data can be Position Embeddings. In addition, it can also include the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user. The above two sub-coding information can be used to characterize the identity of the user. Therefore, the above two sub-coding information can be Identily Embeddings (i.e., an identity embedding vector). The encoding information corresponding to the information interaction sequence data can be as follows Figure 3 As shown, the information interaction sequence data is: [CLS] Can I activate it tomorrow? I don’t want to wait for a month like others [SEP] No [SEP] Be patient, the identification information of the second user is “mentor”, and the E in Token Embeddings [CLS] 、E 能 、E 让 、E 我 、E 明 ……、E 月 、E [SEP] 、E 不 、E 会 ……、E 哦 They represent the embedding vectors of each character, E0, E1, E2, E3, E4, ..., E 30 Respectively represent the encoding information of the position corresponding to each character, E in Identily Embeddings A 、E A 、E A 、E A 、E A ……、E A 、E R 、E R 、E R 、E B 、E B ……、E B Respectively represent the encoding information of different identity information, E A represents the coded information belonging to the first user, E R The coded information representing the identification information of the second user (i.e., “Tutor”), EB Indicates the coded information belonging to the second user.

[0035] In practical applications, the intent recognition model is built by the BERT model.

[0036] Among them, the BERT model is based on the Transformer encoder, and the main model structure is the stacking of Transformers. The corresponding number of latent vectors are obtained through the Transformer encoder of each Transformer layer in the BERT model, and passed to the next Transformer layer, and so on, layer by layer, until the final output result. The model structure of the BERT model for text information classification can be selected, and the BERT model can be constructed through the corresponding algorithm. The corresponding tag sequence placed in each tag position and the historical data corresponding to each tag position can be used to train the above BERT model to obtain the trained BERT model, and finally the intent recognition model can be obtained.

[0037] In actual applications, the information interaction sequence data includes the identification information of the second user set for each second user's interaction information. For example, the information interaction sequence data may be: [CLS] Can I activate it tomorrow? I don't want to wait for a month like others [SEP] Tutor [SEP] No [SEP] Tutor [SEP] Be patient. Correspondingly, TokenEmbeddings, Position Embeddings and Identily Embeddings can be specifically referred to the above processing, such as Figure 4 As shown, Token Embeddings is E [CLS] 、E 能 、E 让 、E 我 、E 明 …、E 月 、E [SEP] 、E 导 、E 师 、E [SEP] 、E 不 、E 会 …、E [SEP] 、E 导 、E 师 、E [SEP] 、E 耐 …、E 哦 , Position Embeddings are E0, E1, E2, E3, E4..., E 36 ,IdentilyEmbeddings is E A 、E A 、E A 、EA 、E A …、E A 、E R 、E R 、E R 、E B 、E B …、E R 、E R 、E R 、E B …、E B .

[0038] Experiments have shown that the accuracy of the intent recognition model can be improved by about 9% after only introducing Identity Embeddings (blocking or hiding the identification information of the second user set by each second user's interaction information in the information interaction sequence data); on the basis of the above, after introducing the identification information of the second user set by each second user's interaction information in the information interaction sequence data, the accuracy of the intent recognition model can be improved by about 12%.

[0039] In actual applications, the information interaction record information between the first user and the second user is located in the complaint voucher image submitted by the first user. Based on this, the information interaction record information between the first user and the second user can be obtained in the following manner. For details, refer to the processing of the following steps A2 to A6.

[0040] In step A2, identification information of a second user who interacts with the first user is identified from the complaint voucher image.

[0041] In implementation, the identification information of the second user who interacts with the first user can be identified from the complaint voucher image through a recognition algorithm or a recognition model, wherein the recognition algorithm can be, for example, an OCR algorithm, and the recognition model can be constructed through a neural network model or a large language model, etc., and can be specifically set according to actual conditions.

[0042] In step A4, information interaction record information belonging to the first user and information interaction record information belonging to the second user are identified from the complaint voucher image.

[0043] In implementation, the information interaction record information belonging to the first user and the information interaction record information belonging to the second user can also be identified from the complaint voucher image through the above-mentioned recognition algorithm or recognition model. For details, please refer to the above content and will not be repeated here.

[0044] In step A6, the identified information is converted into text information, where the text information includes information interaction record information between the first user and the second user.

[0045] In actual applications, the complaint credential image is an image of information interaction between a first user and a second user through an instant messaging application. The second user's identification information includes the nickname information set by the second user in the instant messaging application and / or the note identification information set by the first user for the second user in the instant messaging application. The intention information corresponding to the information interaction record information includes one or more of the transaction reason information of the transaction between the first user and the second user and the occurrence of preset high-risk operations.

[0046] In practical applications, in the above step S108, based on the intention information corresponding to the information interaction record information, the complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. The specific processing method can be varied. The following is another optional processing method, such as Figure 5 As shown, the processing may specifically include the following steps S1082 and S1084.

[0047] In step S1082, when the intention information corresponding to the information interaction record information indicates that there is a preset risk in the transaction between the first user and the second user, a risk prevention and control strategy corresponding to the preset risk is obtained, and the preset risk includes one or more of fraud risk, theft risk, and illegal financial transaction risk.

[0048] In step S1084, based on the acquired risk prevention and control strategy corresponding to the preset risk, risk prevention and control processing is performed on the transaction between the first user and the second user.

[0049] In practical applications, in the above step S108, based on the intention information corresponding to the information interaction record information, the complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. The specific processing method can be varied. The following is another optional processing method, such as Figure 6 As shown, the processing may specifically include the following steps S1086 and S1088.

[0050] In step S1086, when the intention information corresponding to the information interaction record information indicates to remove restrictions on the account of the first user or the second user, a preset account removal strategy is obtained.

[0051] In step S1088, based on the preset account release policy, the account of the first user or the second user is released.

[0052] The embodiment of the present specification provides an information processing method, which receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user. Then, the information interaction record information can be converted into information interaction sequence data that can be input into an intention recognition model. After that, the information interaction sequence data can be input into an encoding module in the intention recognition model. The information interaction sequence data is encoded by the encoding module to obtain encoding information corresponding to the information interaction sequence data. The encoding information corresponding to the information interaction sequence data includes at least sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user. Finally, the encoding information corresponding to the information interaction sequence data can be input into the intention recognition module in the intention recognition model to obtain intent information corresponding to the information interaction record information. The intention information corresponding to the mutually recorded information is determined, and a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. In this way, the identification information of the user in the information interaction process (such as the user's nickname information and / or remark identification information, etc.) is introduced, and the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user are set in the coding information corresponding to the information interaction sequence data, so as to distinguish the subjects in the information interaction (the person himself or the active party, the other party or the passive party), and then restore the information interaction logic, and then perform intent recognition processing, which helps the intent recognition model to better understand the expression content in the information interaction. For the back and forth of information in the information interaction, the scene of repeated mention of keywords can effectively understand and recognize its true intention, improve the understanding of the identities of both parties in the information interaction, and improve the accuracy of intent recognition in the information interaction.

[0053] The above is an information processing method provided in the embodiments of this specification. Based on the same idea, the embodiments of this specification also provide an information processing device, such as Figure 7 shown.

[0054] The information processing device includes: an information receiving module 701, a conversion module 702, an encoding module 703 and an intention determination module 704, wherein: An information receiving module 701 receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user; A conversion module 702 converts the information interaction record information into information interaction sequence data that can be input into an intention recognition model; The encoding module 703 inputs the information interaction sequence data into the encoding module in the intention recognition model, and performs encoding processing on the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user; The intention determination module 704 inputs the coded information corresponding to the information interaction sequence data into the intention recognition module in the intention recognition model, obtains the intention information corresponding to the information interaction record information, and determines the complaint response strategy for the complaint information submitted by the first user based on the intention information corresponding to the information interaction record information to respond to the complaint information of the first user.

[0055] In an embodiment of the present specification, the encoding module includes an encoding module constructed by a Transformer network, and the encoding information corresponding to the information interaction sequence data also includes information interaction sequence sub-coding information corresponding to the information interaction sequence data and sub-coding information of the position corresponding to each character in the information interaction sequence data.

[0056] In the embodiments of this specification, the intent recognition model is constructed by a BERT model.

[0057] In the embodiment of the present specification, the information interaction sequence data includes identification information of the second user set for each interaction information of the second user.

[0058] In the embodiment of this specification, the information interaction record between the first user and the second user is located in the complaint voucher image submitted by the first user, and the device further includes: an identification recognition module, which recognizes identification information of the second user who interacts with the first user from the complaint voucher image; an information identification module, identifying information interaction record information belonging to the first user and information interaction record information belonging to the second user from the complaint voucher image; The text conversion module converts the identified information into text information, wherein the text information includes information interaction record information between the first user and the second user.

[0059] In an embodiment of the present specification, the complaint credential image is an image of information interaction between the first user and the second user through an instant messaging application, the identification information of the second user includes nickname information set by the second user in the instant messaging application and / or remark identification information set by the first user for the second user in the instant messaging application, and the intention information corresponding to the information interaction record information includes one or more of transaction reason information of the transaction between the first user and the second user and the occurrence of preset high-risk operations.

[0060] In the embodiment of this specification, the intention determination module 704 includes: a risk strategy acquisition unit, which, when the intention information corresponding to the information interaction record information indicates that the transaction between the first user and the second user has a preset risk, acquires a risk prevention and control strategy corresponding to the preset risk, wherein the preset risk includes one or more of a fraud risk, a misappropriation risk, and an illegal financial transaction risk; The risk prevention and control unit performs risk prevention and control processing on the transaction between the first user and the second user based on the acquired risk prevention and control strategy corresponding to the preset risk.

[0061] In the embodiment of this specification, the intention determination module 704 includes: a restriction removal strategy acquisition unit, configured to acquire a preset account restriction removal strategy when the intention information corresponding to the information interaction record information indicates that the account of the first user or the second user is to be removed from the restriction; The unblocking unit unblocks the account of the first user or the second user based on a preset account unblocking policy.

[0062] An embodiment of the present specification provides an information processing device, which receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user. Then, the information interaction record information can be converted into information interaction sequence data that can be input into an intention recognition model. Thereafter, the information interaction sequence data can be input into an encoding module in the intention recognition model. The information interaction sequence data is encoded by the encoding module to obtain encoding information corresponding to the information interaction sequence data. The encoding information corresponding to the information interaction sequence data includes at least sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user. Finally, the encoding information corresponding to the information interaction sequence data can be input into the intention recognition module in the intention recognition model to obtain intent information corresponding to the information interaction record information. The intention information corresponding to the mutually recorded information is determined, and a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. In this way, the identification information of the user in the information interaction process (such as the user's nickname information and / or remark identification information, etc.) is introduced, and the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user are set in the coding information corresponding to the information interaction sequence data, so as to distinguish the subjects in the information interaction (the person himself or the active party, the other party or the passive party), and then restore the information interaction logic, and then perform intent recognition processing, which helps the intent recognition model to better understand the expression content in the information interaction. For the back and forth of information in the information interaction, the scene of repeated mention of keywords can effectively understand and recognize its true intention, improve the understanding of the identities of both parties in the information interaction, and improve the accuracy of intent recognition in the information interaction.

[0063] The above is an information processing device provided in the embodiments of this specification. Based on the same idea, the embodiments of this specification also provide an information processing device, such as Figure 8 shown.

[0064] The information processing device may provide a terminal device or a server, etc. for the above-mentioned embodiments.

[0065] The information processing device may have relatively large differences due to different configurations or performances, and may include one or more processors 801 and memory 802, and the memory 802 may store one or more storage applications or data. Among them, the memory 802 may be a short-term storage or a persistent storage. The application stored in the memory 802 may include one or more modules (not shown in the figure), and each module may include a series of computer executable instructions in the information processing device. Furthermore, the processor 801 may be configured to communicate with the memory 802 and execute a series of computer executable instructions in the memory 802 on the information processing device. The information processing device may also include one or more power supplies 803, one or more wired or wireless network interfaces 804, one or more input and output interfaces 805, and one or more keyboards 806.

[0066] Specifically in this embodiment, the information processing device includes a memory and one or more programs, wherein the one or more programs are stored in the memory, and the one or more programs may include one or more modules, and each module may include a series of computer executable instructions for the information processing device, and the one or more programs are configured to be executed by one or more processors, including computer executable instructions for performing the following: receiving complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user; Converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model; Inputting the information interaction sequence data into the encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user; The encoded information corresponding to the information interaction sequence data is input into the intention recognition module in the intention recognition model to obtain the intention information corresponding to the information interaction record information, and based on the intention information corresponding to the information interaction record information, a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user.

[0067] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the information processing device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0068] An embodiment of the present specification provides an information processing device, which receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user. Then, the information interaction record information can be converted into information interaction sequence data that can be input into an intention recognition model. Thereafter, the information interaction sequence data can be input into an encoding module in the intention recognition model. The information interaction sequence data is encoded by the encoding module to obtain encoding information corresponding to the information interaction sequence data. The encoding information corresponding to the information interaction sequence data includes at least sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user. Finally, the encoding information corresponding to the information interaction sequence data can be input into the intention recognition module in the intention recognition model to obtain intent information corresponding to the information interaction record information. The intention information corresponding to the mutually recorded information is determined, and a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. In this way, the identification information of the user in the information interaction process (such as the user's nickname information and / or remark identification information, etc.) is introduced, and the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user are set in the coding information corresponding to the information interaction sequence data, so as to distinguish the subjects in the information interaction (the person himself or the active party, the other party or the passive party), and then restore the information interaction logic, and then perform intent recognition processing, which helps the intent recognition model to better understand the expression content in the information interaction. For the back and forth of information in the information interaction, the scene of repeated mention of keywords can effectively understand and recognize its true intention, improve the understanding of the identities of both parties in the information interaction, and improve the accuracy of intent recognition in the information interaction.

[0069] Furthermore, based on the above Figures 1 to 6 In one embodiment, the present specification further provides a storage medium for storing computer executable instruction information. In a specific embodiment, the storage medium may be a USB flash drive, an optical disk, a hard disk, etc. When the computer executable instruction information stored in the storage medium is executed by the processor, the following process can be implemented: receiving complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user; Converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model; Inputting the information interaction sequence data into the encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user; The encoded information corresponding to the information interaction sequence data is input into the intention recognition module in the intention recognition model to obtain the intention information corresponding to the information interaction record information, and based on the intention information corresponding to the information interaction record information, a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user.

[0070] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the above-mentioned storage medium embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0071] An embodiment of the present specification provides a storage medium, which receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user. Then, the information interaction record information can be converted into information interaction sequence data that can be input into an intention recognition model. Thereafter, the information interaction sequence data can be input into an encoding module in the intention recognition model. The information interaction sequence data is encoded by the encoding module to obtain encoding information corresponding to the information interaction sequence data. The encoding information corresponding to the information interaction sequence data includes at least sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user. Finally, the encoding information corresponding to the information interaction sequence data can be input into the intention recognition module in the intention recognition model to obtain intent information corresponding to the information interaction record information. The intention information corresponding to the information is recorded, and a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. In this way, the identification information of the user in the information interaction process (such as the user's nickname information and / or remark identification information, etc.) is introduced, and the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user are set in the coding information corresponding to the information interaction sequence data, so as to distinguish the subjects in the information interaction (the person himself or the active party, the other party or the passive party), and then restore the information interaction logic, and then perform intent recognition processing, which helps the intent recognition model to better understand the expression content in the information interaction. For the back and forth of information in the information interaction, the scene of repeated mention of keywords can effectively understand and identify its true intention, improve the understanding of the identities of both parties in the information interaction, and improve the accuracy of intent recognition in the information interaction.

[0072] Furthermore, based on the above Figures 1 to 6 In one or more embodiments of the present specification, a computer program product is provided, including a computer program. When the computer program in the computer program product is executed by a processor, the following process can be implemented: receiving complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user; Converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model; Inputting the information interaction sequence data into the encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user; The encoded information corresponding to the information interaction sequence data is input into the intention recognition module in the intention recognition model to obtain the intention information corresponding to the information interaction record information, and based on the intention information corresponding to the information interaction record information, a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user.

[0073] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the above-mentioned computer program product embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0074] The embodiment of the present specification provides a computer program product, which receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user, and then converts the information interaction record information into information interaction sequence data that can be input into an intention recognition model. Thereafter, the information interaction sequence data can be input into an encoding module in the intention recognition model, and the information interaction sequence data is encoded by the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data includes at least sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user. Finally, the encoding information corresponding to the information interaction sequence data can be input into the intention recognition module in the intention recognition model to obtain intent information corresponding to the information interaction record information, and based on the information interaction, the information interaction sequence data can be encoded into the intention recognition module in the intention recognition model to obtain intent information corresponding to the information interaction record information. The intention information corresponding to the mutually recorded information is determined, and a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user. In this way, the identification information of the user in the information interaction process (such as the user's nickname information and / or remark identification information, etc.) is introduced, and the sub-coding information of the user to which each character in the information interaction sequence data belongs and the sub-coding information corresponding to the identification information of the second user set for the interaction information of each second user are set in the coding information corresponding to the information interaction sequence data, so as to distinguish the subjects in the information interaction (the person himself or the active party, the other party or the passive party), and then restore the information interaction logic, and then perform intent recognition processing, which helps the intent recognition model to better understand the expression content in the information interaction. For the back and forth of information in the information interaction, the scene of repeated mention of keywords can effectively understand and recognize its true intention, improve the understanding of the identities of both parties in the information interaction, and improve the accuracy of intent recognition in the information interaction.

[0075] The above is a description of a specific embodiment of the specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0076] In the 1990s, it was very clear whether the improvement of a technology was hardware improvement (for example, improvement of the circuit structure of diodes, transistors, switches, etc.) or software improvement (improvement of the method flow). However, with the development of technology, many improvements of the method flow today can be regarded as direct improvements of the hardware circuit structure. Designers almost always obtain the corresponding hardware circuit structure by programming the improved method flow into the hardware circuit. Therefore, it cannot be said that the improvement of a method flow cannot be implemented with a hardware entity module. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logical function is determined by the user's programming of the device. Designers can "integrate" a digital system on a PLD by programming themselves, without having to ask chip manufacturers to design and make dedicated integrated circuit chips. Moreover, nowadays, instead of manually making integrated circuit chips, this kind of programming is mostly implemented by "logic compiler" software, which is similar to the software compiler used when developing and writing programs, and the original code before compilation must also be written in a specific programming language, which is called hardware description language (HDL). There is not only one kind of HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones are VHDL (Very-High-Speed ​​Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also know that it is only necessary to program the method flow slightly in the above-mentioned hardware description languages ​​and program it into the integrated circuit, and then it is easy to obtain the hardware circuit that implements the logic method flow.

[0077] The controller may be implemented in any suitable manner, for example, the controller may take the form of a microprocessor or processor and a computer-readable medium storing a computer-readable program code (e.g., software or firmware) executable by the (micro)processor, a logic gate, a switch, an application-specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include but are not limited to the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, and the memory controller may also be implemented as part of the control logic of the memory. It is also known to those skilled in the art that, in addition to implementing the controller in a purely computer-readable program code manner, the controller may be implemented in the form of a logic gate, a switch, an application-specific integrated circuit, a programmable logic controller, and an embedded microcontroller by logically programming the method steps. Therefore, such a controller may be considered as a hardware component, and the devices for implementing various functions included therein may also be considered as structures within the hardware component. Or even, the devices for implementing various functions may be considered as both software modules for implementing the method and structures within the hardware component.

[0078] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer may be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.

[0079] For the convenience of description, the above devices are described in terms of functions and are divided into various units. Of course, when implementing one or more embodiments of this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0080] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems, or computer program products. Therefore, one or more embodiments of this specification may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, one or more embodiments of this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0081] The embodiments of this specification are described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of this specification. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable fraud case serial and parallel device to produce a machine, so that the instructions executed by the processor of the computer or other programmable fraud case serial and parallel device generate instructions for implementing the processes in the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0082] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable fraud case serial and parallel device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0083] These computer program instructions may also be loaded onto a computer or other programmable device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0084] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0085] Memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0086] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined in this article, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0087] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0088] Those skilled in the art will appreciate that the embodiments of this specification may be provided as methods, systems or computer program products. Therefore, one or more embodiments of this specification may be in the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Furthermore, one or more embodiments of this specification may be in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0089] One or more embodiments of the present specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. One or more embodiments of the present specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0090] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0091] The above description is only an embodiment of this specification and is not intended to limit this document. For those skilled in the art, this specification may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification should be included in the scope of the claims of this specification.

Claims

1. An information processing method, the method comprising: receiving complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user; Converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model; Inputting the information interaction sequence data into the encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user; The encoded information corresponding to the information interaction sequence data is input into the intention recognition module in the intention recognition model to obtain the intention information corresponding to the information interaction record information, and based on the intention information corresponding to the information interaction record information, a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user.

2. According to the method according to claim 1, the encoding module includes an encoding module constructed by a Transformer network, and the encoding information corresponding to the information interaction sequence data also includes information interaction sequence sub-encoding information corresponding to the information interaction sequence data and sub-encoding information of the position corresponding to each character in the information interaction sequence data.

3. According to the method according to claim 2, the intent recognition model is constructed by a BERT model.

4. The method according to claim 1, wherein the information interaction sequence data includes identification information of the second user set for each interaction information of the second user.

5. According to the method of any one of claims 1 to 4, the information interaction record between the first user and the second user is located in the complaint voucher image submitted by the first user, and the method further comprises: Identifying identification information of the second user who interacts with the first user from the complaint voucher image; Identify information interaction record information belonging to the first user and information interaction record information belonging to the second user from the complaint voucher image; The identified information is converted into text information, where the text information includes information interaction record information between the first user and the second user.

6. According to the method according to claim 5, the complaint credential image is an image of the first user interacting with the second user through an instant messaging application, the identification information of the second user includes nickname information set by the second user in the instant messaging application and / or remark identification information set by the first user for the second user in the instant messaging application, and the intention information corresponding to the information interaction record information includes one or more of transaction reason information of the transaction between the first user and the second user and the occurrence of preset high-risk operations.

7. The method according to claim 6, wherein the step of determining a complaint response strategy for the complaint information submitted by the first user based on the intention information corresponding to the information interaction record information to respond to the complaint information of the first user comprises: When the intention information corresponding to the information interaction record information indicates that there is a preset risk in the transaction between the first user and the second user, obtaining a risk prevention and control strategy corresponding to the preset risk, where the preset risk includes one or more of a fraud risk, a misappropriation risk, and an illegal financial transaction risk; Based on the acquired risk prevention and control strategy corresponding to the preset risk, risk prevention and control processing is performed on the transaction between the first user and the second user.

8. The method according to claim 7, wherein the step of determining a complaint response strategy for the complaint information submitted by the first user based on the intention information corresponding to the information interaction record information to respond to the complaint information of the first user comprises: When the intention information corresponding to the information interaction record information indicates that the account of the first user or the second user is to be unblocked, obtaining a preset account unblocking policy; Based on a preset account unblocking policy, unblock the account of the first user or the second user.

9. An information processing device, comprising: An information receiving module receives complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user; A conversion module converts the information interaction record information into information interaction sequence data that can be input into an intention recognition model; an encoding module, inputting the information interaction sequence data into the encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user; An intention determination module inputs the coded information corresponding to the information interaction sequence data into the intention recognition module in the intention recognition model to obtain the intention information corresponding to the information interaction record information, and based on the intention information corresponding to the information interaction record information, determines a complaint response strategy for the complaint information submitted by the first user to respond to the complaint information of the first user.

10. An information processing device, comprising: processor; as well as a memory arranged to store computer executable instructions which, when executed, cause the processor to: receiving complaint information submitted by a first user, wherein the complaint information includes information interaction record information between the first user and a second user; Converting the information interaction record information into information interaction sequence data that can be input into an intention recognition model; Inputting the information interaction sequence data into the encoding module in the intention recognition model, encoding the information interaction sequence data through the encoding module to obtain encoding information corresponding to the information interaction sequence data, wherein the encoding information corresponding to the information interaction sequence data at least includes sub-coding information of the user to which each character in the information interaction sequence data belongs and sub-coding information corresponding to the identification information of the second user set for each interaction information of the second user; The encoded information corresponding to the information interaction sequence data is input into the intention recognition module in the intention recognition model to obtain the intention information corresponding to the information interaction record information, and based on the intention information corresponding to the information interaction record information, a complaint response strategy for the complaint information submitted by the first user is determined to respond to the complaint information of the first user.

Images