A service provider dialogue processing method and device, a storage medium and an electronic device

Abstract

The specification discloses a service provider dialogue processing method and device, a storage medium and an electronic device, wherein the method comprises: a service provider inputs consultation question data, determines industry identity data corresponding to the service provider, determines consultation request feature data based on the consultation question data and the industry identity data, then determines at least one target knowledge information corresponding to the consultation request feature data based on a knowledge base, and performs dialogue reply processing on the service provider based on the target knowledge information.

Description

Technical Field

[0001] This specification relates to the field of computer technology, and in particular to a service provider dialogue processing method, apparatus, storage medium, and electronic device. Background Technology

[0002] With the development of computer technology and the rapid popularization of electronic devices, various applications and web programs that provide convenient services for life have emerged one after another, providing services for users' food, clothing, housing and transportation (such as travel services, food delivery services, consumer finance services, etc.). Service providers that provide these services to users on service platforms may encounter situations where they need to consult with the service providers when they encounter questions or problems with the use of the services. Summary of the Invention

[0003] This specification provides a service provider dialogue processing method, apparatus, storage medium, and electronic device, the technical solution of which is as follows:

[0004] Firstly, this specification provides a service provider dialogue processing method, the method comprising:

[0005] Obtain the consultation question data input by the service provider, and determine the industry identity data corresponding to the service provider;

[0006] Based on the consultation question data and the industry identity data, consultation request feature data is determined;

[0007] Based on the knowledge base, at least one target knowledge information corresponding to the feature data of the consultation request is determined, and a dialogue response is processed with the service provider based on the target knowledge information.

[0008] Secondly, this specification provides a service provider dialogue processing device, the device comprising:

[0009] The data acquisition module is used to acquire consultation question data input by the service provider and determine the industry identity data corresponding to the service provider;

[0010] The data determination module is used to determine consultation request feature data based on the consultation question data and the industry identity data;

[0011] The dialogue response module is used to determine at least one target knowledge information corresponding to the feature data of the consultation request based on the knowledge base, and to perform dialogue response processing to the service provider based on the target knowledge information.

[0012] Thirdly, this specification provides a computer storage medium storing at least one instruction adapted for loading by a processor and executing method steps of one or more embodiments of this specification.

[0013] Fourthly, this specification provides a computer program product storing at least one instruction adapted to be loaded by a processor and to execute the method steps of one or more embodiments of this specification.

[0014] Fifthly, this specification provides an electronic device that may include: a processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and to execute the method steps of one or more embodiments of this specification.

[0015] The beneficial effects of the technical solutions provided in some embodiments of this specification include at least the following:

[0016] In one or more embodiments of this specification, the service provider inputs consultation question data, determines the industry identity data corresponding to the service provider, determines consultation request feature data based on the consultation question data and industry identity data, then determines at least one target knowledge information corresponding to the consultation request feature data based on the knowledge base, and then performs dialogue response processing to the service provider based on the target knowledge information. This can avoid the situation where the dialogue response obtained by the service provider is often inaccurate due to directly matching the consultation question with the dialogue answer, and can ensure that the knowledge content input to the service provider meets the expectations. It can reduce situations such as consultation questions not matching or irrelevant answers, and improve the accuracy of consultation return matching the correct knowledge of the service provider. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this specification or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of a service provider dialogue processing system provided in this manual;

[0019] Figure 2 This is a flowchart illustrating a service provider dialogue processing method provided in this manual;

[0020] Figure 3 This is a flowchart illustrating another service provider dialogue processing method provided in this manual;

[0021] Figure 4 This is a flowchart illustrating a knowledge base maintenance process provided in this manual;

[0022] Figure 5 This is a schematic diagram of the structure of a service provider dialogue processing device provided in this manual;

[0023] Figure 6 This is a schematic diagram of the structure of a data determination module provided in this specification;

[0024] Figure 7 This is a schematic diagram of the structure of a service provider dialogue processing device provided in this manual;

[0025] Figure 8 This is a schematic diagram of the structure of an electronic device provided in this specification. Detailed Implementation

[0026] The technical solutions in this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this specification.

[0027] In the description of this specification, it should be understood that the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. In the description of this specification, it should be noted that, unless otherwise expressly specified and limited, "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices. Those skilled in the art can understand the specific meaning of the above terms in this specification based on the specific circumstances. Furthermore, in the description of this specification, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, and B alone. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship.

[0028] In related technologies, service providers offering services to users on service platforms often encounter questions or problems regarding service usage, leading to inquiries. Typically, service providers spend considerable time and effort consulting the platform's customer service system via dialogue. However, the responses are often simply matched directly to the question, a common practice that proves inaccurate. For example, a service provider in industry A might obtain knowledge about industry B, which is irrelevant to the service provider in industry A. This demonstrates the limitations of the dialogue processing methods in related technologies.

[0029] The present specification will now be described in detail with reference to specific embodiments.

[0030] Please see Figure 1 This is a schematic diagram of a service provider dialogue processing system provided in this specification. Figure 1 As shown, the service provider dialogue processing system may include at least a client cluster and a service platform 100.

[0031] The client cluster may include at least one client, such as Figure 1 As shown, it specifically includes client 1 corresponding to user 1, client 2 corresponding to user 2, ..., client n corresponding to user n, where n is an integer greater than 0.

[0032] Each client in a client cluster can be an electronic device with communication capabilities, including but not limited to: wearable devices, handheld devices, personal computers, tablets, in-vehicle devices, smartphones, computing devices, or other processing devices connected to a wireless modem. Electronic devices may have different names in different networks, such as: user equipment, access terminal, user unit, user station, mobile station, mobile station, remote station, remote terminal, mobile device, user terminal, terminal, wireless communication equipment, user agent or user device, cellular phone, cordless phone, personal digital assistant (PDA), and electronic devices in 5G networks or future evolved networks.

[0033] The service platform 100 can be a standalone server device, such as a rack-mount, blade, tower, or cabinet-type server device, or a workstation, mainframe, or other hardware device with strong computing power; or it can be a server cluster composed of multiple servers. The servers in the service cluster can be composed in a symmetrical manner, wherein each server is functionally and hierarchically equivalent in the transaction chain, and each server can provide services independently. The independent provision of services can be understood as not requiring the assistance of other servers.

[0034] In one or more embodiments of this specification, the service platform 100 may establish a communication connection with at least one client in the client cluster, and complete the data interaction during the service provider dialogue processing based on the communication connection.

[0035] It should be noted that the service platform 100 establishes a communication connection with at least one client in the client cluster via a network for interactive communication. This network can be a wireless network or a wired network. Wireless networks include, but are not limited to, cellular networks, wireless LANs, infrared networks, or Bluetooth networks. Wired networks include, but are not limited to, Ethernet, universal serial bus (USB), or controller area networks. In one or more embodiments of the specification, technologies and / or formats including Hyper Text Markup Language (HTML), Extensible Markup Language (XML), etc., are used to represent data exchanged over the network (such as target compressed packets). Furthermore, conventional encryption technologies such as Secure Socket Layer (SSL), Transport Layer Security (TLS), Virtual Private Network (VPN), and Internet Protocol Security (IPsec) can be used to encrypt all or some links. In other embodiments, customized and / or dedicated data communication technologies can be used to replace or supplement the aforementioned data communication technologies.

[0036] The service provider dialogue processing system embodiments provided in this specification and the service provider dialogue processing methods described in one or more embodiments belong to the same concept. The execution entity corresponding to the service provider dialogue processing method involved in one or more embodiments of this specification can be the aforementioned service platform 100; the execution entity corresponding to the service provider dialogue processing method involved in one or more embodiments of this specification can also be the electronic device corresponding to the client, specifically determined based on the actual application environment. The implementation process of the service provider dialogue processing system embodiments can be detailed in the following method embodiments, and will not be repeated here.

[0037] based on Figure 1 The following is a detailed description of the service provider dialogue processing method provided by one or more embodiments of this specification, as illustrated in the scenario diagram.

[0038] Please see Figure 2 This document provides a flowchart illustrating a service provider dialogue processing method according to one or more embodiments. This method can be implemented using a computer program and can run on a service provider dialogue processing device based on the von Neumann architecture. The computer program can be integrated into an application or run as a standalone utility application. The service provider dialogue processing device can be a service platform.

[0039] Specifically, the service provider's dialogue handling methods include:

[0040] S102: Obtain the consultation question data input by the service provider and determine the industry identity data corresponding to the service provider;

[0041] In this specification, the scenario involved in dialogue processing can be as follows: taking the knowledge question and answer object service as the intelligent customer service object service as an example, the intelligent customer service object service aims to solve the problem on the device side where the service provider is located. When the service provider asks a question to the intelligent customer service object service, the intelligent customer service object service outputs the answer to the client user.

[0042] For example, a knowledge-based Q&A service could be a chatbot customer service system.

[0043] The consultation question data refers to consultation questions generated by service providers in real-world scenarios based on service usage inquiries or issues.

[0044] For example, a service platform can provide users with lifestyle services, including catering, retail, fast-moving consumer goods, e-commerce, beauty, travel, and rental services. The platform aggregates a large number of service providers, who are the actual providers of these services. These providers can use the platform for operational service management tasks, such as contract signing, merchant expansion, merchant operation, and data viewing. However, service providers may encounter problems and questions at any stage of their operational service management tasks. Therefore, they may trigger the platform's knowledge-based Q&A service through their client application, inputting their questions and concerns into the service.

[0045] In this specification, considering that the dialogue attributes may differ between service providers and ordinary users, if the dialogue processing is modeled after that of ordinary users, the current knowledge return is matched and the answer is returned based on the ordinary user's question itself. This does not take into account the difference between the actual situation when a service provider inquires and the dialogue scenario of an ordinary user, meaning the answers received by the service provider are often inaccurate. For example, a service provider in industry A might receive knowledge about industry B, which would be invalid information for the service provider in industry A. To improve and even resolve this limitation, after obtaining the consultation question data input by the service provider, the service platform first obtains the industry identity data corresponding to the service provider.

[0046] Industry identity data represents the industry attribute data and identity attribute data of the service provider; industry attribute data can be, for example, catering, retail, fast-moving consumer goods, e-commerce, beauty industry, hotel and travel, water and dairy products, etc., and identity attributes can be, for example, merchant user name, merchant position, identity level, etc.

[0047] Understandably, service providers have already registered their relevant business information on the service platform in advance. Based on this, the service platform can obtain the corresponding business information of the service providers, extract data, and obtain industry identity data.

[0048] For example, when a service provider inputs consultation data, the service platform can obtain information such as the current service provider's identity, industry, and name from the service provider user database. At the same time, it can obtain the strategy for the service provider at the time of the consultation from the service strategy database and the set of all platform product capabilities for the service provider from the service product database, thereby obtaining at least the industry identity data.

[0049] S104: Determine consultation request feature data based on the consultation question data and the industry identity data;

[0050] The service platform determines consultation request feature data based on consultation question data and industry identity data. In this specification, knowledge content queries are not performed directly based on consultation question data. Instead, consultation question data and industry identity data are combined to generate consultation request feature data for knowledge content queries.

[0051] Consultation request feature data can integrate merchant consultation information from the question consultation dimension and service provider industry identity information from the service provider characteristic dimension, thereby assisting in the subsequent matching of accurate knowledge content that matches the service provider.

[0052] Optionally, the consultation question data and industry identity data can be concatenated according to a reference data format to obtain consultation request feature data. The reference data format matches the knowledge base data format of the relevant data in the knowledge base, so as to facilitate quick and convenient matching of knowledge content based on the knowledge base in the future.

[0053] Optionally, feature vectors can be extracted from the consultation question data and industry identity data to extract consultation question features and industry identity features. The consultation question features and industry identity features can then be fused to generate comprehensive features, which can be used as consultation request feature data.

[0054] Optionally, consultation request feature data containing multiple reference consultation dimension features can also be determined based on the consultation question data and the industry identity data;

[0055] Among them, the reference consultation dimension features include at least one of the following: service provider name features, consultation question features, service strategy features, service product features, service industry features, and consultation intent features.

[0056] S106: Determine at least one target knowledge information corresponding to the feature data of the consultation request based on the knowledge base, and perform dialogue response processing to the service provider based on the target knowledge information.

[0057] The knowledge base includes multiple reference knowledge feature data, and each reference knowledge feature data contains at least reference knowledge information.

[0058] As an illustration, similar knowledge feature data to the consultation request feature data can be matched in the knowledge base. The feature matching degree between the consultation request feature data and the reference knowledge feature data can be calculated. Based on the feature matching degree, one or more similar knowledge feature data can be filtered out, and then the corresponding target knowledge information can be obtained from the similar knowledge feature data.

[0059] For example, prioritizing reference knowledge feature data based on feature matching degree could involve selecting the number of similar knowledge feature data from targets with high feature matching degree priority.

[0060] In one feasible implementation, the dialogue response processing with the service provider based on the target knowledge information can be carried out in the following manner:

[0061] A2: Obtain the feature matching degree corresponding to one or more target knowledge information;

[0062] A4: If the feature matching degree is greater than the matching degree threshold, then output the target knowledge information with the feature matching degree greater than the matching degree threshold to the service provider;

[0063] The matching degree threshold is a limit or critical value for feature matching degree. When the feature matching degree of the target knowledge information is greater than the matching degree threshold, the accuracy can be considered to be very high and fits the service provider's consultation problem. Conversely, when the feature matching degree of the target knowledge information is less than or equal to the matching degree threshold, the accuracy may generally fit the service provider's consultation problem.

[0064] A6: If the feature matching degree of all target knowledge information is less than the matching degree threshold, then a reference number of recommended knowledge information is determined based on the feature matching degree, a knowledge information set containing the reference number of recommended indication information is generated, and the knowledge information set is output to the service provider in a target query manner.

[0065] Assuming a matching threshold of 80%, if the feature matching degree of the matched target knowledge information does not reach 80%, a reference number (such as the top 5 target knowledge information with the highest feature matching degree) will be selected to combine the knowledge information set. The knowledge information set will be output to the service provider in the form of a target query. After the service provider obtains the knowledge information set, the service provider will determine whether it meets the needs of its consultation question and select the matching target knowledge information that matches the consultation question.

[0066] In one or more embodiments of this specification, the service provider inputs consultation question data, determines the industry identity data corresponding to the service provider, determines consultation request feature data based on the consultation question data and industry identity data, then determines at least one target knowledge information corresponding to the consultation request feature data based on the knowledge base, and then performs dialogue response processing to the service provider based on the target knowledge information. This can avoid the situation where the dialogue response obtained by the service provider is often inaccurate due to directly matching the consultation question with the dialogue answer, and can ensure that the knowledge content input to the service provider meets the expectations. It can reduce situations such as consultation questions not matching or irrelevant answers, and improve the accuracy of consultation return matching the correct knowledge of the service provider.

[0067] Please see Figure 3 , Figure 3 This is a flowchart illustrating another embodiment of a service provider dialogue processing method proposed in one or more embodiments of this specification. Specifically:

[0068] S202: Obtain the consultation question data input by the service provider and determine the industry identity data corresponding to the service provider;

[0069] For details, please refer to the method steps in other embodiments of this specification, which will not be repeated here.

[0070] S204: Based on the consultation question data and the industry identity data, obtain consultation question characteristics, service provider name characteristics, and service industry characteristics;

[0071] Consultation problem characteristics are data features that can reflect the characteristics of consultation problems, determined from the dimensions of consultation problems. Examples include key information of consultation problems, transaction information of consultation problems, metadata information of consultation problems, etc.

[0072] Service provider name features are name data that uniquely identifies a service provider, determined from the service provider dimension.

[0073] Service industry characteristics are feature data that characterizes the industry in which a service provider operates, determined from the perspective of the service industry.

[0074] As an example, the service platform can perform data analysis on consultation question data and industry identity data to extract consultation question characteristics, service provider name characteristics, and service industry characteristics, thereby determining the current service provider's identity, industry, consultation question, and other information.

[0075] S206: Based on the characteristics of the consultation question, the characteristics of the service provider name, the characteristics of the service industry, and the product strategy set, determine the service strategy characteristics, service product characteristics, and consultation intent characteristics;

[0076] A product strategy is a strategic plan developed by a service platform based on the platform's service needs, targeting specific product objects. A product strategy set is a collection of one or more product strategies, and the service platform maintains this product strategy set.

[0077] In this specification, by combining the characteristics of the consultation question, the service provider's name, the service industry, and the current product strategy set of the service platform, we infer the relevant product strategy information, service product information, and consultation intent information of the service provider's current inquiry. Based on the product strategy information, we determine the product strategy characteristics; based on the service product information, we determine the service product characteristics; and based on the consultation intent information, we determine the consultation intent characteristics.

[0078] Service strategy characteristics represent data indicating the actual product strategy corresponding to this service provider's current inquiry.

[0079] Service product characteristics represent the product carrier on which the service provider offers platform services. For example, service product characteristics include, but are not limited to, mini-programs, web pages, applications, etc.

[0080] The characteristics of consultation intent indicate which type of intent the service provider actually has for this consultation, such as consultation for cooperation, consultation for complaints, consultation for appeals, etc.

[0081] For example, if the service provider's name is "Service Provider A", and the service provider A's inquiry question is "What is the Digital Joint Development User Plan?", and the service industry is "water and milk", based on these inquiry question characteristics, service provider name characteristics, and service industry characteristics, and by calling the platform's current product strategy set, the service strategy characteristics of the service provider's actual inquiry are predicted to be "Digital Joint Development Plan", the service product characteristics are "Mini Program", and the inquiry intent characteristics are "cooperative inquiry".

[0082] For example, if the service provider's name is "Service Provider B", and the service provider B's inquiry question is "What are the contents of the mini-program policy?", and the service industry is "logistics", based on these inquiry question characteristics, service provider name characteristics, and service industry characteristics, and by calling the platform's current set of product strategies, it can be predicted that the service strategy characteristics of the service provider's actual inquiry are "mini-program incentive strategy", the service product characteristics are "mini-program", and the inquiry intent characteristics are "appeal".

[0083] In one feasible implementation, a feature matching model can be pre-trained based on a machine learning model, and the feature matching model can be used to associate a set of product strategies for feature processing.

[0084] In practical applications:

[0085] B2: Obtain the feature matching model, which is associated with the product strategy set;

[0086] B2: Input the consultation question characteristics, service provider name characteristics, and service industry characteristics into the feature matching model. Based on the product strategy set, the feature matching model determines the service strategy characteristics, service product characteristics, and consultation intent characteristics for the service provider, and outputs the service strategy characteristics, service product characteristics, and consultation intent characteristics.

[0087] The training process of the feature matching model is explained below:

[0088] Model creation: Create an initial feature matching model for the feature matching scenario based on the machine learning model, and associate the initial feature matching model with the product strategy set;

[0089] Sample data acquisition: Acquire a large amount of sample data. The sample data is based on the historical consultation question data input by the service providers of the service platform. Data processing is performed to extract sample consultation question features, sample service provider name features, and sample service industry features to generate sample data containing sample consultation question features, sample service provider name features, and sample service industry features.

[0090] Sample data annotation: Based on the needs of feature matching scenarios, an expert service is introduced to manually annotate the sample data with corresponding sample tags. The sample tags include service strategy feature tags, service product feature tags, and consultation intent feature tags for each sample data.

[0091] Model training process: Input sample data into the initial feature matching model for at least one round of model training to obtain predicted feature data. The predicted feature data includes predicted service strategy features, predicted service product features, and predicted consultation intent features. Based on the predicted feature data (such as predicted service strategy features, predicted service product features, and predicted consultation intent features) and sample data labels (service strategy feature labels, service product feature labels, and consultation intent feature labels), the model loss function is used to determine the model loss value. Based on the model loss value, the model parameters of the initial feature matching model are adjusted until the model training termination condition is met to obtain the feature matching model.

[0092] Optionally, the model's training termination conditions may include, for example, the loss function value being less than or equal to a preset loss function threshold, or the number of iterations reaching a preset threshold. Specific training termination conditions can be determined based on actual circumstances and are not specifically limited here.

[0093] It should be noted that the machine learning models involved in one or more embodiments of this specification include, but are not limited to, fitting one or more of the following machine learning models: Convolutional Neural Network (CNN) model, Deep Neural Network (DNN) model, Recurrent Neural Networks (RNN) model, embedding model, Gradient Boosting Decision Tree (GBDT) model, Logistic Regression (LR) model, etc.

[0094] Understandably, in practical applications, given the numerous platform products and strategies, training a feature matching model calculates precise service strategy features, service product features, and consultation intent features based on the service provider's inquiry characteristics, service provider name characteristics, and service industry characteristics. This assists in subsequently building high-quality consultation request feature data for knowledge content matching. During model processing, the algorithm can also identify consultation intent, such as consultation intent or complaint intent. Different intent categories will also affect the response method to knowledge feedback during subsequent knowledge base matching. For example, for complaint-type questions, a more gentle and soothing language can be used in the response.

[0095] S208: Based on a preset feature data format, determine consultation request feature data that includes the consultation question features, the service strategy features, the service product features, the service industry features, and the consultation intent features.

[0096] Optionally, service provider name characteristics, consultation question characteristics, service strategy characteristics, service product characteristics, service industry characteristics, and consultation intent characteristics can be concatenated according to a preset feature data format to obtain consultation request feature data. The preset feature data format matches the knowledge base data format of related data in the knowledge base, so as to facilitate quick and convenient matching of knowledge content based on the knowledge base in the future.

[0097] Preset data formats include, but are not limited to, knowledge base data format chains, knowledge base data format tables, knowledge base data vectors, and knowledge base data matrices;

[0098] S210: Determine multiple reference knowledge feature data contained in the knowledge base, wherein the reference knowledge feature data consists of reference consultation feature data and reference knowledge information;

[0099] A knowledge base is pre-built, which includes a large amount of reference knowledge feature data. The reference knowledge feature data consists of reference consultation feature data and reference knowledge information in a certain knowledge base data format.

[0100] For example, the knowledge base data format can be a knowledge base data format table, as shown below:

[0101]

[0102] S212: Determine at least one target knowledge feature data that matches the consultation request feature data from the plurality of reference knowledge feature data;

[0103] Consultation request feature data can be regarded as service provider consultation requests that have been characterized. By matching the consultation request feature data with the reference knowledge feature data in the knowledge base, it is possible to accurately match five dimensions, including service strategy, service product, service industry, consultation intent, and consultation content, thereby obtaining at least one target knowledge feature data.

[0104] In one alternative implementation, the following matching method may be used:

[0105] C2: The service platform determines the feature matching degree between the reference consultation feature data and the consultation request feature data based on the reference knowledge feature data and the reference consultation feature data.

[0106] Data similarity algorithms can be used to calculate the feature matching degree between each reference consultation feature data and the consultation request feature data. Data similarity algorithms include, but are not limited to, the TF-IDF (term frequency–inverse document frequency) algorithm and the calculation of vector distance between data (distance can include, but is not limited to, cosine distance, Euclidean distance, Manhattan distance, Mahalanobis distance, or Minkowski distance, etc.).

[0107] In a typical example, feature engineering is used to vectorize the reference consultation feature data and consultation request feature data into reference consultation feature vectors and consultation request feature vectors. Then, the feature vector distance between the two feature vectors is calculated (the distance can include, but is not limited to, cosine distance, Euclidean distance, Manhattan distance, Mahalanobis distance, or Minkowski distance, etc.). The feature vector distance is used as the feature matching degree.

[0108] C4: Based on the feature matching degree, determine at least one target knowledge feature data that matches the consultation request feature data from the plurality of reference knowledge feature data.

[0109] In a demonstrative sense, similar knowledge feature data to the consultation request feature data can be matched in the knowledge base. The feature matching degree between the consultation request feature data and the reference knowledge feature data can be calculated. Based on the feature matching degree, one or more similar knowledge feature data can be selected as target knowledge feature data. Then, the corresponding target knowledge information can be obtained from the target knowledge feature data.

[0110] For example, prioritizing reference knowledge feature data based on feature matching degree can be achieved by selecting the number of similar knowledge feature data with high feature matching degree priority as the target knowledge feature data.

[0111] S214: Obtain at least one target knowledge information from the target knowledge feature data.

[0112] Since the target knowledge feature data consists of consultation feature data and knowledge information, the knowledge information in the target knowledge feature data can be obtained as the target knowledge information.

[0113] In one or more embodiments of this specification, the above method is used to reconstruct consultation questions to obtain consultation request feature data, and then perform subsequent matching based on the consultation request feature data. This can avoid the situation where directly matching dialogue answers according to consultation questions often results in inaccurate dialogue answers obtained by service providers. It can ensure that the knowledge content input to service providers meets expectations, reduce situations such as consultation questions not matching or answers not being asked, and improve the accuracy of consultation returns matching the correct knowledge of service providers.

[0114] Please see Figure 4 , Figure 4 This is a schematic diagram of a knowledge base maintenance process proposed in one or more embodiments of this specification. Specifically:

[0115] S302: Determine the target number of sample knowledge feature data, wherein the sample knowledge feature data consists of sample consultation question features, sample service strategy features, sample service product features, sample service industry features, sample consultation intent features, and sample knowledge information;

[0116] To reduce the difficulty of knowledge base maintenance and improve processing efficiency, the process of selecting tags such as industry and service strategy for knowledge entries can be simplified each time data is entered into the knowledge base. Only a small number of sample knowledge feature data points need to be prepared in advance, for example, 500 knowledge points, and stored in the knowledge base. Using this small amount of sample knowledge feature data, a machine learning-based knowledge feature model can be trained. This model can then automatically improve the consultation feature data based on key knowledge information.

[0117] Understandably, after training, the knowledge feature model takes key knowledge information as input and automatically traverses all or part of the databases associated with the service platform (such as service strategy database, promotion plan database, service product database, etc.) to extract knowledge features and then generates reference consultation feature data.

[0118] For example, sample knowledge feature data can be used as sample data labels for knowledge feature models during the model training phase;

[0119] S304: Based on the sample knowledge feature data, train the initial knowledge feature model corresponding to the knowledge base to obtain the trained knowledge feature model;

[0120] For example, the following explains the model training process of a knowledge feature model:

[0121] Model creation: Creating an initial knowledge feature model based on a machine learning model;

[0122] Training data preparation: The target number of sample knowledge feature data is used as the label for the model training data in the model training stage; the model training data is the key information of sample knowledge, that is, a large amount of key information of sample knowledge is collected, which can be extracted from the sample knowledge feature data.

[0123] Among them, the (sample) knowledge key data can be data of key characteristic types such as knowledge title, knowledge summary, and key knowledge points.

[0124] Model training process:

[0125] During the forward training of the model, key information of sample knowledge is input into the initial knowledge feature model for at least one round of model training to obtain predicted knowledge feature data. The predicted knowledge feature data consists of predicted consulting question features, predicted service strategy features, predicted service product features, predicted service industry features, predicted consulting intent features, and predicted knowledge information.

[0126] During the reverse training of the model, the key information prediction loss is calculated based on the predicted knowledge feature data and the sample knowledge feature data. The model parameters of the initial knowledge feature model are adjusted based on the key information prediction loss until the initial knowledge feature model meets the end training conditions, and the trained knowledge feature model is obtained.

[0127] During the model's backpropagation training, the gradient direction of the knowledge feature model is determined by predicting the loss based on key information using the backpropagation algorithm. The model parameters in the knowledge feature model are then updated layer by layer from the output layer forward, thus correcting the model parameters and obtaining the corrected knowledge feature model.

[0128] The key information prediction loss can be obtained using model loss calculation formulas in related technologies, such as cross-entropy loss, Euclidean distance loss, etc. That is, the first loss is calculated based on the predicted consulting question features and the sample consulting question features, the second loss is calculated based on the predicted service strategy features and the sample service strategy features, the third loss is calculated based on the predicted service product features and the sample service product features, the fourth loss is calculated based on the predicted service industry features and the sample service industry features, the fifth loss is calculated based on the predicted consulting intention features and the sample consulting intention features, and the sixth loss is calculated based on the predicted knowledge information and the sample knowledge information. The key information prediction loss is obtained based on the first, second, third, fourth, fifth, and sixth losses.

[0129] This knowledge feature model has the ability to extract event content features, summarize event content semantic knowledge, and learn and memorize information features during the semantic extraction and generation process. Typically, the information or knowledge learned by the machine learning model is stored in the connection matrix between each unit node.

[0130] Model application: Deploy the trained knowledge feature model to the knowledge base maintenance scenario for processing.

[0131] Optionally, the machine learning model can be implemented by fitting one or more of the following models: Convolutional Neural Network (CNN), Deep Neural Network (DNN), Recurrent Neural Networks (RNN), Embedding model, Gradient Boosting Decision Tree (GBDT) model, Logistic Regression (LR) model, etc.

[0132] S306: Obtain key knowledge information for reference knowledge information, determine reference consultation feature data based on the key knowledge information using a knowledge feature model, and save the reference consultation feature data to the knowledge base.

[0133] In one feasible implementation, the service platform inputs key knowledge information into a knowledge feature model, and uses the knowledge feature model to determine features of reference consultation questions, reference service strategies, reference service products, reference service industries, and reference consultation intentions, as well as reference knowledge information. The platform then outputs reference consultation feature data containing these features, along with reference consultation feature information.

[0134] In one or more embodiments of this specification, introducing a knowledge feature model in the above manner can reduce the difficulty of the knowledge base maintenance phase, simplify the difficulty of manually selecting tags such as industry and service strategy of knowledge each time knowledge base data is entered, optimize the dialogue processing flow, and reduce the work of operation personnel in entering information such as product, strategy, industry and intent type of knowledge through the model processing of the knowledge feature model, thereby improving the efficiency of knowledge entry, especially for a large amount of knowledge, improving processing efficiency.

[0135] The following will combine Figure 5 This manual provides a detailed description of the service provider dialogue processing device provided. It should be noted that... Figure 5 The service provider dialogue processing device shown is used to execute this specification. Figures 1-6 The methods of the embodiments shown are illustrated only in connection with this specification for ease of explanation. For specific technical details not disclosed, please refer to this specification. Figures 1-4 The example shown.

[0136] Please see Figure 5 This diagram illustrates the structure of the service provider dialogue processing device 1 described in this specification. This service provider dialogue processing device 1 can be implemented as all or part of a user terminal through software, hardware, or a combination of both. According to some embodiments, the service provider dialogue processing device 1 includes a data acquisition module 11, a data determination module 12, and a dialogue response module 13, specifically used for:

[0137] Data acquisition module 11 is used to acquire consultation question data input by the service provider and determine the industry identity data corresponding to the service provider;

[0138] Data determination module 12 is used to determine consultation request feature data based on the consultation question data and the industry identity data;

[0139] The dialogue response module 13 is used to determine at least one target knowledge information corresponding to the feature data of the consultation request based on the knowledge base, and to perform dialogue response processing to the service provider based on the target knowledge information.

[0140] Optionally, the data determination module 12 is used for:

[0141] Based on the consultation question data and the industry identity data, consultation request feature data containing multiple reference consultation dimension features is determined;

[0142] The reference consultation dimension features include at least one of the following: service provider name features, consultation question features, service strategy features, service product features, service industry features, and consultation intent features.

[0143] Optional, such as Figure 6As shown, the data determination module 12 includes:

[0144] Feature acquisition unit 121 is used to acquire consultation question features, service provider name features and service industry features based on the consultation question data and the industry identity data;

[0145] The feature determination unit 122 is used to determine service strategy features, service product features, and consultation intent features based on the consultation question features, service provider name features, service industry features, and product strategy set.

[0146] The data determination unit 123 is used to determine consultation request feature data, which includes the consultation question features, the service strategy features, the service product features, the service industry features, and the consultation intent features, based on a preset feature data format.

[0147] Optionally, the feature determination unit 122 is used for:

[0148] Obtain a feature matching model, which is associated with a set of product strategies;

[0149] The consultation question characteristics, service provider name characteristics, and service industry characteristics are input into the feature matching model. Based on the product strategy set, the feature matching model determines the service strategy characteristics, service product characteristics, and consultation intent characteristics for the service provider, and outputs the service strategy characteristics, the service product characteristics, and the consultation intent characteristics.

[0150] Optionally, the dialogue response module 13 is used for:

[0151] The knowledge base contains multiple reference knowledge feature data, which consists of reference consultation feature data and reference knowledge information;

[0152] Determine at least one target knowledge feature data that matches the consultation request feature data from the plurality of reference knowledge feature data;

[0153] At least one target knowledge information is obtained from the target knowledge feature data.

[0154] Optionally, the dialogue response module 13 is used for:

[0155] Based on the reference consultation feature data corresponding to the reference knowledge feature data, determine the feature matching degree between the reference consultation feature data and the consultation request feature data;

[0156] Based on the feature matching degree, at least one target knowledge feature data that matches the consultation request feature data is determined from the plurality of reference knowledge feature data.

[0157] Optional, such as Figure 7 As shown, the device 1 further includes:

[0158] The knowledge base maintenance module 14 is used to determine the target number of sample knowledge feature data, which consists of sample consultation question features, sample service strategy features, sample service product features, sample service industry features, sample consultation intent features, and sample knowledge information.

[0159] The knowledge base maintenance module 14 is used to train the initial knowledge feature model corresponding to the knowledge base based on the sample knowledge feature data to obtain the trained knowledge feature model.

[0160] The knowledge base maintenance module 14 is used to acquire key knowledge information for reference knowledge information, determine reference consultation feature data based on the key knowledge information using a knowledge feature model, and save the reference consultation feature data to the knowledge base.

[0161] Optionally, the knowledge base maintenance module 14 is used for

[0162] The key knowledge information is input into the knowledge feature model, which determines the features of the reference consultation question, the reference service strategy, the reference service product, the reference service industry, the reference consultation intent, and the reference knowledge information. The output is reference consultation feature data containing the features of the reference consultation question, the reference service strategy, the reference service product, the reference service industry, the reference consultation intent, and the reference knowledge information.

[0163] Optionally, the dialogue response module is used for:

[0164] Obtain the feature matching degree corresponding to the target knowledge information;

[0165] If the feature matching degree is greater than the matching degree threshold, then the target knowledge information is output to the service provider;

[0166] If the feature matching degree of all the target knowledge information is less than the matching degree threshold, then a reference number of recommended knowledge information is determined based on the feature matching degree, a knowledge information set containing the reference number of recommended indication information is generated, and the knowledge information set is output to the service provider in a target query manner.

[0167] It should be noted that the service provider dialogue processing device provided in the above embodiments is only illustrated by the division of the above functional modules when executing the service provider dialogue processing method. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the service provider dialogue processing device and the service provider dialogue processing method embodiments provided in the above embodiments belong to the same concept, and the implementation process is detailed in the method embodiments, which will not be repeated here.

[0168] The serial numbers in this specification are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0169] This specification also provides a computer storage medium capable of storing multiple instructions adapted to be loaded and executed by a processor as described above. Figures 1-4 The service provider dialogue processing method described in the illustrated embodiment can be found in the following documentation for its specific execution process. Figures 1-4 The specific details of the illustrated embodiments will not be elaborated here.

[0170] This specification also provides a computer program product that stores at least one instruction, said at least one instruction being loaded and executed by the processor as described above. Figures 1-4 The service provider dialogue processing method described in the illustrated embodiment can be found in the following documentation for its specific execution process. Figures 1-4 The specific details of the illustrated embodiments will not be elaborated here.

[0171] Please refer to Figure 8 This is a structural block diagram of an electronic device provided in an embodiment of this specification. The electronic device in this specification may include one or more of the following components: a processor 110, a memory 120, an input device 130, an output device 140, and a bus 150. The processor 110, memory 120, input device 130, and output device 140 can be connected via the bus 150.

[0172] Processor 110 may include one or more processing cores. Processor 110 connects to various parts of the terminal using various interfaces and lines, and performs various functions and processes data of terminal 100 by running or executing instructions, programs, code sets, or instruction sets stored in memory 120, and by calling data stored in memory 120. Optionally, processor 110 may be implemented using at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), or programmable logic array (PLA). Processor 110 may integrate one or more of the following: central processing unit (CPU), graphics processing unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem handles wireless communication. It is understood that the modem may also not be integrated into processor 110 and may be implemented separately using a communication chip.

[0173] The memory 120 may include random access memory (RAM) or read-only memory (ROM). Optionally, the memory 120 may include non-transitory computer-readable storage medium. The memory 120 may be used to store instructions, programs, code, code sets, or instruction sets.

[0174] The input device 130 is used to receive input instructions or data, and includes, but is not limited to, a keyboard, mouse, camera, microphone, or touch device. The output device 140 is used to output instructions or data, and includes, but is not limited to, a display device and a speaker. In this embodiment, the input device 130 can be a temperature sensor to obtain the operating temperature of the terminal. The output device 140 can be a speaker to output audio signals.

[0175] In addition, those skilled in the art will understand that the structure of the terminal shown in the above figures does not constitute a limitation on the terminal. The terminal may include more or fewer components than shown, or combine certain components, or have different component arrangements. For example, the terminal may also include radio frequency circuits, input units, sensors, audio circuits, wireless fidelity (WIFI) modules, power supplies, Bluetooth modules, etc., which will not be described in detail here.

[0176] In the embodiments of this specification, the executing entity for each step can be the terminal described above. Optionally, the executing entity for each step is the terminal's operating system. The operating system can be Android, iOS, or other operating systems; this specification does not limit this.

[0177] exist Figure 8 In the electronic device, the processor 110 can be used to call a program stored in the memory 120 and execute it to implement the service provider dialogue processing method as described in the various method embodiments of this specification.

[0178] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory, or random access memory, etc.

[0179] It should be noted that the information (including but not limited to user device information, user personal information, etc.), data (including but not limited to data used for analysis, stored data, displayed data, etc.), and signals involved in the embodiments of this specification are all authorized by the user or fully authorized by all parties, and the collection, use, and processing of related data must comply with the relevant laws, regulations, and standards of the relevant countries and regions. For example, the object characteristics, interactive behavior characteristics, and user information involved in this specification were all obtained under full authorization.

[0180] The above-disclosed embodiments are merely preferred embodiments of this specification and should not be construed as limiting the scope of this specification. Therefore, any equivalent variations made in accordance with the claims of this specification shall still fall within the scope of this specification.

Claims

1. A service provider dialogue processing method, the method comprising: Obtain the consultation question data input by the service provider, and determine the industry identity data corresponding to the service provider; Based on the consultation question data and the industry identity data, we obtain consultation question characteristics, service provider name characteristics, and service industry characteristics; Based on the characteristics of the consultation question, the characteristics of the service provider name, the characteristics of the service industry, and the product strategy set, the characteristics of the service strategy, the characteristics of the service product, and the characteristics of the consultation intent are determined. Based on a preset feature data format, consultation request feature data is determined, which includes the features of the consultation question, the service strategy, the service product, the service industry, and the consultation intent. Based on the knowledge base, at least one target knowledge information corresponding to the feature data of the consultation request is determined, and a dialogue response is processed to the service provider based on the target knowledge information. The step of determining at least one target knowledge information corresponding to the feature data of the consultation request based on the knowledge base includes: The knowledge base contains multiple reference knowledge feature data, which consists of reference consultation feature data and reference knowledge information. The consultation request feature data is matched with the reference knowledge feature data. The data matching dimensions include service strategy, service product, service industry, consultation intent, and consultation content, so as to determine at least one target knowledge feature data that matches the consultation request feature data from the multiple reference knowledge feature data, and to obtain at least one target knowledge information from the target knowledge feature data.

2. The method according to claim 1, wherein determining the service strategy characteristics, service product characteristics, and consultation intent characteristics based on the consultation question characteristics, service provider name characteristics, service industry characteristics, and product strategy set includes: Obtain a feature matching model, which is associated with a set of product strategies; The consultation question characteristics, service provider name characteristics, and service industry characteristics are input into the feature matching model. Based on the product strategy set, the feature matching model determines the service strategy characteristics, service product characteristics, and consultation intent characteristics for the service provider, and outputs the service strategy characteristics, the service product characteristics, and the consultation intent characteristics.

3. The method according to claim 1, wherein determining at least one target knowledge feature data matching the consultation request feature data from the plurality of reference knowledge feature data comprises: Based on the reference consultation feature data corresponding to the reference knowledge feature data, determine the feature matching degree between the reference consultation feature data and the consultation request feature data; Based on the feature matching degree, at least one target knowledge feature data that matches the consultation request feature data is determined from the plurality of reference knowledge feature data.

4. The method according to claim 1, further comprising: The target number of sample knowledge feature data is determined. The sample knowledge feature data consists of sample consultation question features, sample service strategy features, sample service product features, sample service industry features, sample consultation intent features, and sample knowledge information. Based on the sample knowledge feature data, the initial knowledge feature model corresponding to the knowledge base is trained to obtain the trained knowledge feature model. Obtain key knowledge information for reference knowledge information, determine reference consultation feature data based on the key knowledge information using a knowledge feature model, and save the reference consultation feature data to the knowledge base.

5. The method according to claim 4, wherein determining the reference consultation feature data based on the key knowledge information using a knowledge feature model includes: The key knowledge information is input into the knowledge feature model, and then... The knowledge feature model identifies the features of reference consultation questions, reference service strategies, reference service products, reference service industries, and reference consultation intentions, as well as reference knowledge information. It outputs reference consultation feature data containing these features.

6. The method according to claim 1, wherein the dialogue response processing based on the target knowledge information for the service provider includes: Obtain the feature matching degree corresponding to the target knowledge information; If the feature matching degree is greater than the matching degree threshold, then the target knowledge information is output to the service provider; If the feature matching degree of all the target knowledge information is less than the matching degree threshold, then a reference number of recommended knowledge information is determined based on the feature matching degree, a knowledge information set containing the reference number of recommended indication information is generated, and the knowledge information set is output to the service provider in a target query manner.

7. A service provider dialogue processing device, the device comprising: The data acquisition module is used to acquire consultation question data input by the service provider and determine the industry identity data corresponding to the service provider; The data determination module is used to obtain consultation question characteristics, service provider name characteristics, and service industry characteristics based on the consultation question data and the industry identity data; Based on the consultation question characteristics, service provider name characteristics, service industry characteristics, and product strategy set, service strategy characteristics, service product characteristics, and consultation intent characteristics are determined. Based on a preset feature data format, consultation request feature data containing the consultation question characteristics, service strategy characteristics, service product characteristics, service industry characteristics, and consultation intent characteristics is determined. The dialogue response module is used to determine at least one target knowledge information corresponding to the feature data of the consultation request based on the knowledge base, and to perform dialogue response processing to the service provider based on the target knowledge information. The step of determining at least one target knowledge information corresponding to the feature data of the consultation request based on the knowledge base includes: The knowledge base contains multiple reference knowledge feature data, which consists of reference consultation feature data and reference knowledge information. The consultation request feature data is matched with the reference knowledge feature data. The data matching dimensions include service strategy, service product, service industry, consultation intent, and consultation content, so as to determine at least one target knowledge feature data that matches the consultation request feature data from the multiple reference knowledge feature data, and to obtain at least one target knowledge information from the target knowledge feature data.

8. A computer storage medium storing a plurality of instructions adapted for loading by a processor and executing the method steps of any one of claims 1 to 6.

9. A computer program product storing at least one instruction, said at least one instruction being loaded by a processor and executing the method steps of any one of claims 1 to 6.

10. An electronic device, comprising: A processor and a memory; wherein the memory stores a computer program adapted to be loaded by the processor and executed the method steps as claimed in any one of claims 1 to 6.

Images