Replay processing method and device based on artificial intelligence, computer equipment and medium

Abstract

The invention belongs to the technical field of artificial intelligence, and relates to an artificial intelligence-based reply processing method and device, computer equipment and a storage medium, and the method comprises the steps: obtaining question information inputted by a target user from a user session corresponding to the target user; preprocessing the problem information to obtain target problem information; calling a problem identification model to carry out problem identification processing on the target problem information to obtain a corresponding problem identification result; wherein the problem identification result comprises a target problem type and a target problem reason corresponding to the target problem information; obtaining a target reply template corresponding to the target question type; generating an initial reply based on the target question reason and a target reply template; performing adjustment processing on the initial reply based on a preset adjustment strategy to obtain a target reply; and returning the target reply to the target user. In addition, the target reply can be stored in the block chain. According to the method, the accuracy of reply processing is effectively improved, and the insurance efficiency and satisfaction of the user are improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence development technology and the field of financial technology, and in particular to artificial intelligence-based reply processing methods, devices, computer equipment and storage media. Background Art

[0002] In the current digital age, with the rapid development of mobile Internet technology, more and more users tend to use smartphone applications (APP) to complete various services, including car insurance purchases. The car insurance purchase platform developed by insurance companies provides users with a convenient channel for car insurance purchases. Its main process includes key steps such as vehicle entry, quotation information entry, and confirmation of insurance plans.

[0003] However, in the actual insurance process, users often encounter various problems, resulting in interruptions in the insurance process. These problems include but are not limited to inconsistent information between the person and the car, inconsistent ID numbers, etc. When encountering these problems, users usually choose to go to the online customer service for answers. However, the existing online customer service system can often only give general replies based on the error messages that appear during the auto insurance purchase process, and cannot provide targeted guidance for specific problems encountered by users. This response method not only reduces the accuracy of the response, but may also cause users to communicate with customer service multiple times to solve the problem, thereby reducing the processing efficiency of auto insurance. Summary of the invention

[0004] The purpose of the embodiments of the present application is to propose an artificial intelligence-based reply processing method, device, computer equipment and storage medium to solve the technical problem that the existing online customer service system can only give general replies based on the error information that appears during the purchase of auto insurance, and cannot provide targeted guidance for the specific problems encountered by the user, thereby reducing the accuracy of the reply and reducing the processing efficiency of auto insurance.

[0005] In order to solve the above technical problems, the embodiment of the present application provides a reply processing method based on artificial intelligence, which adopts the following technical solutions:

[0006] Acquire question information input by the target user from a user session corresponding to the target user;

[0007] Preprocessing the problem information to obtain corresponding target problem information;

[0008] Calling a pre-trained problem recognition model to perform problem recognition processing on the target problem information to obtain a problem recognition result corresponding to the target problem information; wherein the problem recognition result includes a target problem type and a target problem cause corresponding to the target problem information;

[0009] Obtaining a target response template corresponding to the target question type;

[0010] Generate a corresponding initial response based on the target problem cause and the target response template;

[0011] Adjusting the initial response based on a preset adjustment strategy to obtain a corresponding target response;

[0012] The target reply is returned to the target user.

[0013] Furthermore, the step of obtaining a target response template corresponding to the target question type specifically includes:

[0014] Call the preset reply template library;

[0015] Filtering a first response template matching the target question type from the response template library;

[0016] Determining whether there is more than one first reply template;

[0017] If so, obtaining reply preference information corresponding to the target user;

[0018] A second reply template corresponding to the reply preference information is screened out from all the first reply templates, and the second reply template is used as the target reply template.

[0019] Furthermore, the step of generating a corresponding initial response based on the target problem cause and the target response template specifically includes:

[0020] identifying all placeholders from the target response template;

[0021] Identifying a designated placeholder corresponding to the target problem cause from all the placeholders, and performing data replacement on the designated placeholder based on the target problem cause;

[0022] Collect the specified data corresponding to other placeholders from the specified data source;

[0023] Perform corresponding data replacement processing on the other placeholders based on the specified data to obtain a processed target reply template;

[0024] The processed target response template is used as the initial response.

[0025] Furthermore, the step of adjusting the initial response based on a preset adjustment strategy to obtain a corresponding target response specifically includes:

[0026] Performing grammatical adjustment processing on the initial reply to obtain a corresponding first reply;

[0027] Performing spelling adjustment on the first reply to obtain a corresponding second reply;

[0028] Performing readability adjustment processing on the second reply to obtain a corresponding third reply;

[0029] The third reply is used as the target reply.

[0030] Furthermore, the step of returning the target reply to the target user specifically includes:

[0031] Get multiple preset push channels;

[0032] Filter out target push channels from all the push channels;

[0033] The target reply is returned to the target user through the target push channel.

[0034] Furthermore, the step of obtaining corresponding target question information based on preprocessing the question information specifically includes:

[0035] Remove stop words from the question information to obtain corresponding first question information;

[0036] Performing word segmentation processing on the first question information to obtain corresponding second question information;

[0037] Performing format conversion processing on the second question to obtain corresponding third question information;

[0038] The third question information is used as the target question information.

[0039] Further, after the step of returning the target reply to the target user, the method further includes:

[0040] Acquire historical behavior information corresponding to the target user;

[0041] Constructing a database query statement based on the problem information and the historical behavior information;

[0042] Execute the database query statement to query the associated questions and answers corresponding to the question information from a preset question database;

[0043] Encapsulating the associated question and the reply into corresponding response objects;

[0044] Push the response object to the target user.

[0045] In order to solve the above technical problems, the embodiment of the present application further provides a reply processing device based on artificial intelligence, which adopts the following technical solution:

[0046] A first acquisition module, configured to acquire question information input by a target user from a user session corresponding to the target user;

[0047] A preprocessing module, used to preprocess the problem information to obtain corresponding target problem information;

[0048] An identification module, used to call a pre-trained problem identification model to perform problem identification processing on the target problem information, and obtain a problem identification result corresponding to the target problem information; wherein the problem identification result includes a target problem type and a target problem cause corresponding to the target problem information;

[0049] A second acquisition module is used to acquire a target response template corresponding to the target question type;

[0050] A generating module, configured to generate a corresponding initial response based on the target problem cause and the target response template;

[0051] An adjustment module, used to adjust the initial response based on a preset adjustment strategy to obtain a corresponding target response;

[0052] A returning module is used to return the target reply to the target user.

[0053] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0054] Acquire question information input by the target user from a user session corresponding to the target user;

[0055] Preprocessing the problem information to obtain corresponding target problem information;

[0056] Calling a pre-trained problem recognition model to perform problem recognition processing on the target problem information to obtain a problem recognition result corresponding to the target problem information; wherein the problem recognition result includes a target problem type and a target problem cause corresponding to the target problem information;

[0057] Obtaining a target response template corresponding to the target question type;

[0058] Generate a corresponding initial response based on the target problem cause and the target response template;

[0059] Adjusting the initial response based on a preset adjustment strategy to obtain a corresponding target response;

[0060] The target reply is returned to the target user.

[0061] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0062] Acquire question information input by the target user from a user session corresponding to the target user;

[0063] Preprocessing the problem information to obtain corresponding target problem information;

[0064] Calling a pre-trained problem recognition model to perform problem recognition processing on the target problem information to obtain a problem recognition result corresponding to the target problem information; wherein the problem recognition result includes a target problem type and a target problem cause corresponding to the target problem information;

[0065] Obtaining a target response template corresponding to the target question type;

[0066] Generate a corresponding initial response based on the target problem cause and the target response template;

[0067] Adjusting the initial response based on a preset adjustment strategy to obtain a corresponding target response;

[0068] The target reply is returned to the target user.

[0069] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0070] The present application first obtains the question information input by the target user from the user session corresponding to the target user; then pre-processes the question information to obtain the corresponding target question information; then calls the pre-trained question recognition model to perform question recognition processing on the target question information to obtain a question recognition result corresponding to the target question information; wherein the question recognition result includes a target question type and a target question cause corresponding to the target question information; subsequently obtains a target reply template corresponding to the target question type; and generates a corresponding initial reply based on the target question cause and the target reply template; further adjusts the initial reply based on a preset adjustment strategy to obtain a corresponding target reply; and finally returns the target reply to the target user. The present application obtains question information input by the target user from the user session corresponding to the target user, and pre-processes the question information to obtain the target question information, and then based on the use of a pre-trained question recognition model, it can quickly and accurately obtain the target question type and target question cause corresponding to the target question information, and then obtain a target reply template corresponding to the target question type, and then generate a corresponding initial reply based on the target question cause and the target reply template, and automatically and intelligently adjust the initial reply based on the adjustment strategy to obtain the target reply and return it to the target user, thereby realizing automatic and accurate processing of user question replies during the auto insurance underwriting process, improving the accuracy of reply processing, and thus helping to improve user insurance efficiency and satisfaction. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0072] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0073] Figure 2 A flowchart of an embodiment of an artificial intelligence-based reply processing method according to the present application;

[0074] Figure 3 is a structural schematic diagram of an embodiment of an artificial intelligence-based reply processing device according to the present application;

[0075] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0076] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0077] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0078] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0079] like Figure 1 As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0080] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0081] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV), a laptop computer, a desktop computer, etc.

[0082] The server 103 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal device 101 .

[0083] It should be noted that the artificial intelligence-based reply processing method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the artificial intelligence-based reply processing device is generally set in the server / terminal device.

[0084] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to the implementation requirements.

[0085] Continue to refer Figure 2 , showing a flow chart of an embodiment of an artificial intelligence-based reply processing method according to the present application. According to different requirements, the order of the steps in the flow chart can be changed, and some steps can be omitted. The artificial intelligence-based reply processing method provided in the embodiment of the present application can be applied to any scenario where reply processing is required, and the artificial intelligence-based reply processing method can be applied to products in these scenarios, for example, reply processing in the financial insurance field. The artificial intelligence-based reply processing method comprises the following steps:

[0086] Step S201: obtaining question information input by a target user from a user session corresponding to the target user.

[0087] In this embodiment, the electronic device (e.g. Figure 1The server / terminal device shown in the figure) can obtain the question information input by the target user through a wired connection or a wireless connection. It should be pointed out that the above-mentioned wireless connection method may include but is not limited to 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or to be developed in the future. The question information is the problem information in the process. The executor of the present application may be a car insurance customer service system, or simply referred to as a system, which is a system with an intelligent reply function. The present application can be applied to business scenarios in which automated and intelligent reply processing is performed on questions submitted by users during the car insurance process. Among them, the above-mentioned question information may be the process problem information entered by the target user in the user session during the self-service car insurance process, such as inconsistent person-vehicle information, inconsistent certificate numbers, and failure to transmit vehicle transfer signs. The above-mentioned user session may include chat windows, telephone calls, etc.

[0088] Step S202: pre-process the question information to obtain corresponding target question information.

[0089] In this embodiment, the above-mentioned specific implementation process of preprocessing the question information to obtain the corresponding target question information will be further described in detail in subsequent specific embodiments of the present application, and will not be elaborated in detail here.

[0090] Step S203, calling a pre-trained problem recognition model to perform problem recognition processing on the target problem information to obtain a problem recognition result corresponding to the target problem information; wherein the problem recognition result includes a target problem type and a target problem cause corresponding to the target problem information.

[0091] In this embodiment, the process of building a problem identification model includes: collecting a large amount of user problem data from multiple channels (such as customer service records, user feedback, online forums, etc.) in advance. Then clean the collected data to remove duplicate, irrelevant or low-quality questions. After that, manually annotate the cleaned data and assign one or more category labels to each question (such as incorrect vehicle information, quotation problems, payment problems, etc.). And further annotate the specific causes of the problem (such as inconsistent certificate numbers, inaccurate vehicle transfer marks, etc.). Subsequently, use the annotated data set to train a custom model so that it can accurately classify user inputs and output the corresponding problem types and causes of the problems. Among them, the custom model can be a classifier model trained using a deep learning framework (such as TensorFlow, PyTorch, etc.).

[0092] The application process of the question recognition model includes: when the user inputs a question, it is first preprocessed, such as removing stop words, word segmentation, removing punctuation marks, etc. Then the preprocessed input question is passed to the trained question recognition model, and the question recognition model will output one or more category labels as the prediction result. Specifically, according to the output of the question recognition model, a set threshold (such as a probability threshold) is used to determine which category the input question belongs to. If the model outputs multiple category labels, the label with the highest probability can be selected as the final classification. In addition, if the question recognition model can identify specific errors or problems (this usually requires the inclusion of more fine-grained labels when training the model), the specific cause of the input problem, that is, the cause of the problem, is further determined based on the output of the model.

[0093] Step S204, obtaining a target response template corresponding to the target question type.

[0094] In this embodiment, the specific implementation process of obtaining the target response template corresponding to the target question type will be described in further detail in subsequent specific embodiments of the present application and will not be elaborated on here.

[0095] Step S205: Generate a corresponding initial response based on the target problem cause and the target response template.

[0096] In this embodiment, the specific implementation process of generating the corresponding initial response based on the target problem cause and the target response template will be described in further detail in subsequent specific embodiments of the present application and will not be elaborated on here.

[0097] Step S206: adjusting the initial response based on a preset adjustment strategy to obtain a corresponding target response.

[0098] In this embodiment, the above-mentioned specific implementation process of adjusting the initial response based on the preset adjustment strategy to obtain the corresponding target response will be further described in detail in subsequent specific embodiments of the present application, and will not be elaborated in detail here.

[0099] Step S207: returning the target reply to the target user.

[0100] In this embodiment, the specific implementation process of returning the target reply to the target user will be described in further detail in subsequent specific embodiments of the present application, and will not be elaborated on here.

[0101] The present application first obtains the question information input by the target user from the user session corresponding to the target user; then pre-processes the question information to obtain the corresponding target question information; then calls the pre-trained question recognition model to perform question recognition processing on the target question information to obtain a question recognition result corresponding to the target question information; wherein the question recognition result includes a target question type and a target question cause corresponding to the target question information; subsequently obtains a target reply template corresponding to the target question type; and generates a corresponding initial reply based on the target question cause and the target reply template; further adjusts the initial reply based on a preset adjustment strategy to obtain a corresponding target reply; and finally returns the target reply to the target user. The present application obtains question information input by the target user from the user session corresponding to the target user, and pre-processes the question information to obtain the target question information, and then based on the use of a pre-trained question recognition model, it can quickly and accurately obtain the target question type and target question cause corresponding to the target question information, and then obtain a target reply template corresponding to the target question type, and then generate a corresponding initial reply based on the target question cause and the target reply template, and automatically and intelligently adjust the initial reply based on the adjustment strategy to obtain the target reply and return it to the target user, thereby realizing automatic and accurate processing of user question replies during the auto insurance underwriting process, improving the accuracy of reply processing, and thus helping to improve user insurance efficiency and satisfaction.

[0102] In some optional implementations, step S204 includes the following steps:

[0103] Call the preset reply template library.

[0104] In this embodiment, a reply template library is pre-designed according to actual business needs, and the reply template library contains multiple types of questions and their corresponding reply templates. Each reply template should contain a standard answer format for a specific question type and a placeholder for inserting dynamic data.

[0105] Among them, for each step of the user's auto insurance operation (including process information, page information, car information, insurance agency, insurance channel, etc.), error information (error steps, error reasons), questions asked after entering the customer service and replies are synchronized to big data. Big data analyzes and classifies these data, and seeks a complete set of reply templates for each of these questions. In addition, the user information, car information and insurance information involved in the auto insurance process can be saved in the blockchain, which can ensure that the data is not changed, so that big data can obtain accurate data when analyzing the data, thereby improving the accuracy of the analysis.

[0106] A first response template matching the target question type is selected from the response template library.

[0107] In this embodiment, a first reply template matching the generated target question type can be found in the reply template library. The number of first reply templates may include one or more. Specifically, the same question category may have multiple variant templates.

[0108] Determine whether there is a plurality of the first reply templates.

[0109] In this embodiment, the number of first reply templates can be counted to determine whether the number of the first reply templates is one or more. If the number of the first reply template is one, the first reply template is directly used as the target reply template.

[0110] If so, obtain reply preference information corresponding to the target user.

[0111] In this embodiment, the reply preference information refers to the template information preferred by the target user.

[0112] A second reply template corresponding to the reply preference information is screened out from all the first reply templates, and the second reply template is used as the target reply template.

[0113] In this embodiment, if there are multiple first reply templates matching the target question type, a second reply template that meets the user's reply preference can be screened out from all the first reply templates and used as the target reply template.

[0114] In addition, the most appropriate reply template may be selected from all first reply templates based on specific details of the problem (such as the severity and urgency of the problem). The selection of the template may also be adjusted based on the user's historical behavior or preference (such as whether the user has previously expressed satisfaction or dissatisfaction with a certain type of reply).

[0115] This application calls a preset reply template library; then filters out a first reply template that matches the target question type from the reply template library; then determines whether there are multiple first reply templates; if so, obtains reply preference information corresponding to the target user; subsequently filters out a second reply template corresponding to the reply preference information from all the first reply templates, and uses the second reply template as the target reply template. This application filters out a first reply template that matches the target question type based on the use of a reply template library. If it is detected that there are multiple first reply templates, it intelligently obtains the reply preference information corresponding to the target user, and then filters out a second reply template corresponding to the reply preference information from all the first reply templates, and uses the second reply template as the target reply template, effectively ensuring the intelligence and accuracy of the filtered target reply templates.

[0116] In some optional implementations of this embodiment, step S205 includes the following steps:

[0117] All placeholders are identified from the target reply template.

[0118] In this embodiment, all placeholders included in the target reply template can be identified by performing placeholder identification on the target reply template, wherein the placeholder is usually marked as a specific string (such as {{user_name}}, {{vehicle_id}}, etc.) to indicate the location where dynamic data needs to be inserted.

[0119] A designated placeholder corresponding to the target problem cause is identified from all the placeholders, and data of the designated placeholder is replaced based on the target problem cause.

[0120] In this embodiment, the placeholders include a designated placeholder for inserting the problem reason. The target problem reason can be replaced according to the position of the designated placeholder, and a string replacement operation can be used to replace the designated placeholder with the corresponding data value (target problem reason). In the data replacement process, the accuracy and integrity of the data need to be ensured to avoid generating incorrect reply content.

[0121] Collect designated data corresponding to other placeholders from a designated data source; wherein the other placeholders are placeholders other than the designated placeholder among all the placeholders.

[0122] In this embodiment, the above-mentioned designated data sources may at least include user input, system database, third-party services, etc. According to the identified other placeholders, the required data can be collected from the above-mentioned designated data sources, and the collected data needs to be properly processed (such as formatting, cleaning, conversion, etc.) to ensure that they are compatible with the placeholders in the target reply template, so as to obtain the corresponding designated data.

[0123] Based on the specified data, corresponding data replacement processing is performed on the other placeholders to obtain a processed target reply template.

[0124] In this embodiment, the above other placeholders can be replaced with corresponding specified data. During the replacement process, the accuracy and completeness of the specified data are ensured to avoid generating erroneous reply content. After the data replacement process is completed, the placeholders in the target reply template will be replaced by the actual data values, thereby generating the final reply content, that is, the target reply template after the above processing.

[0125] The processed target response template is used as the initial response.

[0126] The present application identifies all placeholders from the target reply template; then identifies the designated placeholder corresponding to the target problem cause from all the placeholders, and replaces the data of the designated placeholder based on the target problem cause; then collects designated data corresponding to other placeholders from a designated data source; subsequently performs corresponding data replacement processing on the other placeholders based on the designated data to obtain a processed target reply template; and uses the processed target reply template as the initial reply. The present application identifies all placeholders from the target reply template, and then performs data replacement processing on the placeholders in the target reply template based on the target problem cause and the use of designated data to generate a final initial reply, which effectively ensures the accuracy and consistency of the initial reply and helps to improve user experience and satisfaction.

[0127] In some optional implementations, step S206 includes the following steps:

[0128] The initial reply is subjected to grammatical adjustment processing to obtain a corresponding first reply.

[0129] In this embodiment, the initial reply can be grammatically analyzed by using NLP tools, specifically by identifying sentence structure, subject-verb-object relationships, etc., and pointing out any possible grammatical errors, and then correcting the grammatical errors in the initial reply to obtain the corresponding first reply.

[0130] The first reply is spelled and adjusted to obtain a corresponding second reply.

[0131] In this embodiment, the spelling errors in the first reply can be checked by using a spell checker with a spell checking function. Among them, a dictionary containing industry terms, common abbreviations and specific words that may be entered by the user can be maintained. The dictionary is then used to help the spell checker identify and correctly process these words to avoid misjudging them as spelling errors. In addition, spell checking can give priority to contextual information. For example, if a specific brand or place name is mentioned in the reply, and the name does not exist in the general dictionary, but it can be judged that it is correct based on the context, then the spell checker should be able to recognize and ignore this "error".

[0132] The second reply is processed for readability to obtain a corresponding third reply.

[0133] In this embodiment, the readability adjustment process may include: checking whether the sentences in the second reply are too long or too complex. If the sentence is too long, the second reply can be split into shorter sentences; if the sentence is too complex, it can be simplified. And ensure that there is a good paragraph structure in the content of the second reply, and each paragraph revolves around a central idea, which helps users better understand and absorb information. In addition, the language style of the reply can be adjusted according to the language preferences and habits of the target user group. For example, for young user groups, a more colloquial and relaxed language style can be used; for professional user groups, a more formal and professional language style may need to be used.

[0134] The third reply is used as the target reply.

[0135] The present application performs grammatical adjustment processing on the initial reply to obtain the corresponding first reply; then performs spelling adjustment processing on the first reply to obtain the corresponding second reply; then performs readability adjustment processing on the second reply to obtain the corresponding third reply; and subsequently uses the third reply as the target reply. The present application performs grammatical adjustment processing, spelling adjustment processing, and readability adjustment processing on the initial reply, so as to quickly and accurately complete the adjustment processing of the initial reply and obtain the corresponding target reply, effectively ensuring the accuracy and readability of the generated target reply. This allows the target reply to be returned to the target user in the future, which is conducive to improving the user experience of the target user.

[0136] In some optional implementations, step S207 includes the following steps:

[0137] Get multiple preset push channels.

[0138] In this embodiment, the above-mentioned push channels may at least include system chat windows, emails, text messages and other channels.

[0139] Screen out the target push channel from all the said push channels.

[0140] In this embodiment, one channel can be randomly screened out from the above-mentioned push channels as the above-mentioned target push channel.

[0141] Return the target reply to the target user through the target push channel.

[0142] In this embodiment, before returning the target reply to the target user, the target reply can also be further customized or optimized as needed to meet the specific needs or preferences of the target user.

[0143] This application obtains a variety of preset push channels; then screens out the target push channel from all the said push channels; subsequently, returns the target reply to the target user through the target push channel. By screening out the target push channel from a variety of preset push channels and then returning the target reply to the target user through this target push channel, it is beneficial to improve the intelligence of the target reply push.

[0144] In some optional implementation manners of this embodiment, step S202 includes the following steps:

[0145] Perform stop-word removal processing on the problem information to obtain the corresponding first problem information.

[0146] In this embodiment, the above-mentioned stop-word removal processing may refer to removing stop words in the problem information, such as meaningless words like "de" (的) and "le" (了).

[0147] Perform word segmentation processing on the first problem information to obtain the corresponding second problem information.

[0148] In this embodiment, the above-mentioned word segmentation processing refers to splitting the sentences in the first problem information into meaningful word units.

[0149] Perform format conversion processing on the second problem to obtain the corresponding third problem information.

[0150] In this embodiment, the above-mentioned format conversion processing refers to converting the second problem into a format that can be processed by the problem recognition model, such as vector representation.

[0151] Use the third problem information as the target problem information.

[0152] The present application obtains the corresponding first question information by removing stop words from the question information; then performs word segmentation on the first question information to obtain the corresponding second question information; then performs format conversion on the second question to obtain the corresponding third question information; and subsequently uses the third question information as the target question information. The present application removes stop words, performs word segmentation, and performs format conversion on the question information, thereby achieving fast and accurate preprocessing of the question information and ensuring the data accuracy of the generated target question information.

[0153] In some optional implementations of this embodiment, after step S207, the electronic device may further perform the following steps:

[0154] Obtain historical behavior information corresponding to the target user.

[0155] In this embodiment, historical behavior information corresponding to the target user can be queried from a preset behavior database based on the user information of the target user. The behavior database is a pre-built database storing behavior information of multiple users' problem queries in the auto insurance self-service process.

[0156] A database query statement is constructed based on the problem information and the historical behavior information.

[0157] In this embodiment, the above database query statement is used to search for related questions and replies related to the user's question (ie, question information) in the question database.

[0158] The database query statement is executed to query the associated questions and answers corresponding to the question information from a preset question database.

[0159] In this embodiment, by executing the above database query statement, the related questions and replies related to the question information are searched from the question database. After that, the question database returns the query result, including the title, description and corresponding reply of the related question.

[0160] The associated question and the reply are encapsulated into corresponding response objects.

[0161] In this embodiment, the associated questions and the replies may be packaged into a response object, which includes a list of associated questions and the reply to each question.

[0162] Push the response object to the target user.

[0163] In this embodiment, the response object can be pushed to the target user by using a push channel such as a chat window, email, or text message.

[0164] This application obtains the historical behavior information corresponding to the target user; then constructs a database query statement based on the question information and the historical behavior information; then executes the database query statement to query the associated questions and replies corresponding to the question information from the preset question database; subsequently encapsulates the associated questions and replies into corresponding response objects, and pushes the response objects to the target user. This application intelligently obtains the historical behavior information corresponding to the target user after returning the target reply to the target user, and constructs a database query statement based on the question information and the historical behavior information, and then executes the database query statement to query the associated questions and replies corresponding to the question information from the question database, and subsequently pushes the response objects encapsulated by the associated questions and replies to the target user, so as to automatically and intelligently provide the target user with associated questions and replies related to the current question, which is conducive to improving the target user's satisfaction and service quality.

[0165] In some optional implementations, the user information obtained is subject to the user's consent and complies with relevant laws and policies.

[0166] In addition, the non-Company software tools or components that appear in the embodiments of the present application are merely examples and do not represent actual use.

[0167] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0168] It should be emphasized that in order to further ensure the privacy and security of the above-mentioned target response, the above-mentioned target response can also be stored in a node of a blockchain.

[0169] The blockchain referred to in this application is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0170] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0171] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0172] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0173] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0174] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of a reply processing device based on artificial intelligence, and the device embodiment is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0175] like Figure 3As shown, the artificial intelligence-based reply processing device 300 described in this embodiment includes: a first acquisition module 301, a preprocessing module 302, a recognition module 303, a second acquisition module 304, a generation module 305, an adjustment module 306 and a return module 307. Among them:

[0176] A first acquisition module 301 is used to acquire question information input by a target user from a user session corresponding to the target user;

[0177] A preprocessing module 302 is used to preprocess the question information to obtain corresponding target question information;

[0178] The identification module 303 is used to call the pre-trained problem identification model to perform problem identification processing on the target problem information to obtain a problem identification result corresponding to the target problem information; wherein the problem identification result includes a target problem type and a target problem cause corresponding to the target problem information;

[0179] A second acquisition module 304 is used to acquire a target response template corresponding to the target question type;

[0180] A generating module 305, configured to generate a corresponding initial response based on the target problem cause and the target response template;

[0181] An adjustment module 306, configured to adjust the initial response based on a preset adjustment strategy to obtain a corresponding target response;

[0182] The return module 307 is used to return the target reply to the target user.

[0183] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based reply processing method in the aforementioned implementation mode, and will not be repeated here.

[0184] In some optional implementations of this embodiment, the second acquisition module 304 includes:

[0185] The calling submodule is used to call the preset reply template library;

[0186] A first screening submodule, configured to screen out a first reply template matching the target question type from the reply template library;

[0187] A judging submodule, used for judging whether the number of the first reply templates is multiple;

[0188] A first acquisition submodule, configured to acquire reply preference information corresponding to the target user;

[0189] The second screening submodule is used to screen out a second reply template corresponding to the reply preference information from all the first reply templates, and use the second reply template as the target reply template.

[0190] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based reply processing method in the aforementioned implementation mode, and will not be repeated here.

[0191] In some optional implementations of this embodiment, the generation module 305 includes:

[0192] An identification submodule, used to identify all placeholders from the target reply template;

[0193] A first replacement submodule, configured to identify a designated placeholder corresponding to the target problem cause from all the placeholders, and replace data of the designated placeholder based on the target problem cause;

[0194] A collection submodule, used to collect specified data corresponding to other placeholders from a specified data source; wherein the other placeholders are placeholders other than the specified placeholder among all the placeholders;

[0195] A second replacement submodule is used to perform corresponding data replacement processing on the other placeholders based on the specified data to obtain a processed target reply template;

[0196] The first determining submodule is configured to use the processed target reply template as the initial reply.

[0197] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based reply processing method in the aforementioned implementation mode, and will not be repeated here.

[0198] In some optional implementations of this embodiment, the adjustment module 306 includes:

[0199] A first adjustment submodule, configured to perform grammatical adjustment processing on the initial reply to obtain a corresponding first reply;

[0200] A second adjustment submodule, configured to perform spelling adjustment processing on the first reply to obtain a corresponding second reply;

[0201] A third adjustment submodule, configured to perform readability adjustment processing on the second reply to obtain a corresponding third reply;

[0202] The second determining submodule is configured to use the third reply as the target reply.

[0203] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based reply processing method in the aforementioned implementation mode, and will not be repeated here.

[0204] In some optional implementations of this embodiment, the returning module 307 includes:

[0205] The second acquisition submodule is used to acquire multiple preset push channels;

[0206] A third screening submodule is used to screen out a target push channel from all the push channels;

[0207] The return submodule is used to return the target reply to the target user through the target push channel.

[0208] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based reply processing method in the aforementioned implementation mode, and will not be repeated here.

[0209] In some optional implementations of this embodiment, the preprocessing module 302 includes:

[0210] A first processing submodule, configured to remove stop words from the question information to obtain corresponding first question information;

[0211] A second processing submodule is used to perform word segmentation processing on the first question information to obtain corresponding second question information;

[0212] A third processing submodule is used to perform format conversion processing on the second question to obtain corresponding third question information;

[0213] The third determining submodule is used to use the third question information as the target question information.

[0214] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based reply processing method in the aforementioned implementation mode, and will not be repeated here.

[0215] In some optional implementations of this embodiment, the reply processing device based on artificial intelligence further includes:

[0216] A third acquisition module is used to acquire historical behavior information corresponding to the target user;

[0217] A construction module, used to construct a database query statement based on the problem information and the historical behavior information;

[0218] A query module, used to execute the database query statement to query the associated questions and answers corresponding to the question information from a preset question database;

[0219] An encapsulation module, used for encapsulating the associated question and the reply into a corresponding response object;

[0220] A push module is used to push the response object to the target user.

[0221] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based reply processing method in the aforementioned implementation mode, and will not be repeated here.

[0222] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0223] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0224] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0225] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 4. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the reply processing method based on artificial intelligence, etc. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0226] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run computer-readable instructions stored in the memory 41 or process data, such as computer-readable instructions for running the artificial intelligence-based reply processing method.

[0227] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0228] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0229] In an embodiment of the present application, question information input by the target user is obtained from a user session corresponding to the target user, and the question information is pre-processed to obtain the target question information. Then, based on the use of a pre-trained question recognition model, the target question type and the target question cause corresponding to the target question information can be quickly and accurately obtained. Subsequently, a target reply template corresponding to the target question type is obtained, and then a corresponding initial reply is generated based on the target question cause and the target reply template. The initial reply is automatically and intelligently adjusted based on the adjustment strategy to obtain the target reply and return it to the target user, thereby realizing automatic and accurate question reply processing for users during the auto insurance underwriting process, improving the accuracy of reply processing, and thus facilitating improving user insurance efficiency and satisfaction.

[0230] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the above-mentioned artificial intelligence-based reply processing method.

[0231] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0232] In an embodiment of the present application, question information input by the target user is obtained from a user session corresponding to the target user, and the question information is pre-processed to obtain the target question information. Then, based on the use of a pre-trained question recognition model, the target question type and the target question cause corresponding to the target question information can be quickly and accurately obtained. Subsequently, a target reply template corresponding to the target question type is obtained, and then a corresponding initial reply is generated based on the target question cause and the target reply template. The initial reply is automatically and intelligently adjusted based on the adjustment strategy to obtain the target reply and return it to the target user, thereby realizing automatic and accurate question reply processing for users during the auto insurance underwriting process, improving the accuracy of reply processing, and thus facilitating improving user insurance efficiency and satisfaction.

[0233] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0234] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A reply processing method based on artificial intelligence, characterized in that: The steps include: Acquire question information input by the target user from a user session corresponding to the target user; Preprocessing the problem information to obtain corresponding target problem information; Calling a pre-trained problem recognition model to perform problem recognition processing on the target problem information to obtain a problem recognition result corresponding to the target problem information; wherein the problem recognition result includes a target problem type and a target problem cause corresponding to the target problem information; Obtaining a target response template corresponding to the target question type; Generate a corresponding initial response based on the target problem cause and the target response template; Adjusting the initial response based on a preset adjustment strategy to obtain a corresponding target response; The target reply is returned to the target user.

2. The artificial intelligence-based reply processing method according to claim 1, characterized in that: The step of obtaining a target response template corresponding to the target question type specifically includes: Call the preset reply template library; Filtering a first response template matching the target question type from the response template library; Determining whether there is more than one first reply template; If so, obtaining reply preference information corresponding to the target user; A second reply template corresponding to the reply preference information is screened out from all the first reply templates, and the second reply template is used as the target reply template.

3. The artificial intelligence-based reply processing method according to claim 1, characterized in that: The step of generating a corresponding initial response based on the target problem cause and the target response template specifically includes: identifying all placeholders from the target response template; Identifying a designated placeholder corresponding to the target problem cause from all the placeholders, and performing data replacement on the designated placeholder based on the target problem cause; Collecting specified data corresponding to other placeholders from a specified data source; wherein the other placeholders are placeholders other than the specified placeholder among all the placeholders; Perform corresponding data replacement processing on the other placeholders based on the specified data to obtain a processed target reply template; The processed target response template is used as the initial response.

4. The artificial intelligence-based reply processing method according to claim 1, characterized in that: The step of adjusting the initial response based on a preset adjustment strategy to obtain a corresponding target response specifically includes: Performing grammatical adjustment processing on the initial reply to obtain a corresponding first reply; Performing spelling adjustment on the first reply to obtain a corresponding second reply; Performing readability adjustment processing on the second reply to obtain a corresponding third reply; The third reply is used as the target reply.

5. The artificial intelligence-based reply processing method according to claim 1, characterized in that: The step of returning the target reply to the target user specifically includes: Get multiple preset push channels; Filter out target push channels from all the push channels; The target reply is returned to the target user through the target push channel.

6. The artificial intelligence-based reply processing method according to claim 1, characterized in that: The step of obtaining corresponding target question information based on preprocessing the question information specifically includes: Remove stop words from the question information to obtain corresponding first question information; Performing word segmentation processing on the first question information to obtain corresponding second question information; Performing format conversion processing on the second question to obtain corresponding third question information; The third question information is used as the target question information.

7. The artificial intelligence-based reply processing method according to claim 1, characterized in that: After the step of returning the target reply to the target user, the method further includes: Acquire historical behavior information corresponding to the target user; Constructing a database query statement based on the problem information and the historical behavior information; Execute the database query statement to query the associated questions and answers corresponding to the question information from a preset question database; Encapsulating the associated question and the reply into corresponding response objects; Push the response object to the target user.

8. A reply processing device based on artificial intelligence, characterized in that: include: A first acquisition module, configured to acquire question information input by a target user from a user session corresponding to the target user; A preprocessing module, used to preprocess the problem information to obtain corresponding target problem information; An identification module, used to call a pre-trained problem identification model to perform problem identification processing on the target problem information, and obtain a problem identification result corresponding to the target problem information; wherein the problem identification result includes a target problem type and a target problem cause corresponding to the target problem information; A second acquisition module is used to acquire a target response template corresponding to the target question type; A generating module, configured to generate a corresponding initial response based on the target problem cause and the target response template; An adjustment module, used to adjust the initial response based on a preset adjustment strategy to obtain a corresponding target response; A returning module is used to return the target reply to the target user.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the reply processing method based on artificial intelligence as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps of the artificial intelligence-based reply processing method as described in any one of claims 1 to 7 are implemented.

Images