Response method and device based on user objection, electronic equipment and storage medium
By obtaining user information, building user portraits and using objection response generation models, it solves the problem that customer service personnel find it difficult to deal with customer personalized objections, and improves the recommendation success rate of insurance products.
Patent Information
- Application Number
- CN202411266729.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-10
- Publication Date
- 2025-01-03
AI Technical Summary
When promoting insurance products, it is difficult for customer service personnel to quickly adjust communication methods and service strategies to deal with customer personalized objections, resulting in a decrease in the recommendation rate of insurance products.
By obtaining the information of the target user, building a user portrait, and using preset objection response generation models, intention recognition and emotion recognition technologies, personalized response text is generated to deal with customer objections.
It improves the recommendation success rate of insurance products, and adjusts service strategies to match customers' specific needs and expectations through in-depth understanding of customers' needs and emotions.
Smart Images
Figure CN119248925A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of text generation, and in particular to a method and device for responding to user objections, an electronic device, and a storage medium. Background Art
[0002] When customer service staff are rejected when selling insurance products to customers, it is difficult for them to quickly adjust their communication methods and service strategies to address their objections, which leads to a decrease in the recommendation rate of insurance products. Therefore, how to improve the success rate of insurance product recommendations has become an urgent problem to be solved. Summary of the invention
[0003] The main purpose of the embodiments of the present application is to propose a method and device, an electronic device and a storage medium for responding to user objections, aiming to improve the success rate of recommending insurance products.
[0004] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present application proposes a method for responding to user objections, the method comprising:
[0005] Acquire target user information of a target object, acquire target product information of a target product, and acquire target objection information of the target object to the target product;
[0006] Building a target user portrait according to the target user information;
[0007] Performing intent recognition on the target objection information to obtain target intent information;
[0008] Performing emotion recognition on the target objection information to obtain target emotion information;
[0009] Generate an objection response using a preset objection response generation model, the target user portrait, the target intent information, and the target product information to obtain an initial response text;
[0010] Respond based on the target emotion information and the initial response text.
[0011] In some embodiments, the target user information includes: age data, gender data, geographic location data, purchase history data, historical browsing data, occupation data, and income data; and constructing a target user portrait based on the target user information includes:
[0012] Constructing a user identity profile based on the age data, gender data and geographic location data;
[0013] Building a purchase tendency profile based on the purchase history data and the historical browsing data;
[0014] constructing a purchasing power profile based on the occupation data and the income data;
[0015] The target user portrait is obtained by performing portrait aggregation based on the user identity portrait, the purchase tendency portrait and the purchasing power portrait.
[0016] In some embodiments, constructing a purchase tendency profile based on the purchase history data and the historical browsing data includes:
[0017] Perform product preference identification based on the purchase history data and the historical browsing data to obtain product preference information;
[0018] Performing time series analysis based on the purchase history data and the historical browsing data to obtain product purchase trends;
[0019] The purchase tendency portrait is constructed according to the product purchase trend and the product preference information.
[0020] In some embodiments, responding based on the target emotion information and the initial response text includes:
[0021] Extracting emotional keywords from the initial response text to obtain initial emotional keywords;
[0022] Performing vocabulary adjustment on the initial emotion keywords according to the target emotion information to obtain target tone keywords;
[0023] Performing text adjustment on the initial response text based on the target tone keywords to obtain a target response text;
[0024] A response is made based on the target response text.
[0025] In some embodiments, the performing intent recognition on the target objection information to obtain the target intent information includes:
[0026] Performing semantic recognition on the target objection information to obtain target objection semantics;
[0027] Classifying the target objection semantics according to a preset objection category to obtain a target objection category;
[0028] Intent recognition is performed based on the target objection semantics and the target objection category to obtain the target intent information.
[0029] In some embodiments, performing emotion recognition on the target objection information to obtain target emotion information includes:
[0030] Performing keyword extraction on the target objection information to obtain target keyword information;
[0031] According to a preset emotional keyword dictionary, the target keyword information is subjected to emotional tendency recognition to obtain the keyword emotional tendency;
[0032] The keyword sentiment tendency is weightedly calculated to obtain the target sentiment information.
[0033] In some embodiments, the objection response generation model includes: an encoding module, an attention module, a normalization module and a residual block; the objection response generation model is preset, the target user portrait, the target intent information and the target product information are used to generate the response to obtain the initial response text, including:
[0034] Performing text encoding on the target user portrait, the target intent information and the target product information through the encoding module to obtain text encoding information;
[0035] Extracting features of the text encoding information through the attention module to obtain text feature information;
[0036] Performing feature normalization on the text feature information by the normalization module to obtain text normalization information;
[0037] The text normalization information is subjected to residual connection through the residual block to obtain the initial response text.
[0038] To achieve the above-mentioned purpose, a second aspect of an embodiment of the present application proposes a response device based on user objection, the device comprising:
[0039] An information acquisition module is used to acquire target user information of a target object, acquire target product information of a target product, and acquire target objection information of the target object to the target product;
[0040] A portrait building module, used to build a target user portrait according to the target user information;
[0041] An intention recognition module, used to perform intention recognition on the target objection information to obtain target intention information;
[0042] An emotion recognition module, used to perform emotion recognition on the target objection information to obtain target emotion information;
[0043] A response generation module, used to generate a response through a preset objection response generation model, the target user portrait, the target intent information and the target product information to obtain an initial response text;
[0044] A text response module is used to respond based on the target emotional information and the initial response text.
[0045] To achieve the above objectives, a third aspect of an embodiment of the present application proposes an electronic device, which includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the method described in the first aspect is implemented.
[0046] To achieve the above-mentioned purpose, the fourth aspect of an embodiment of the present application proposes a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method described in the first aspect is implemented.
[0047] The present application proposes a method and device for responding to user objections, an electronic device, and a storage medium. The method obtains target user information of a target object, target product information of a target product, and target objection information, then constructs a target user portrait based on the target user information, and performs intent recognition on the target objection information. The method then generates a response using a preset objection response generation model, the target user portrait, the target intent information, and the target product information, thereby generating an initial response text with personalized responses to products and users, and achieving rapid adjustment of service strategies in response to customer objections. At the same time, the method performs emotion recognition on the target objection information to obtain target emotion information, and finally responds using the target emotion information and the initial text, thereby improving the success rate of product recommendations. BRIEF DESCRIPTION OF THE DRAWINGS
[0048] Figure 1 is a flow chart of a method for responding to user objections provided in an embodiment of the present application;
[0049] Figure 2 yes Figure 1 Flow chart of step S102 in FIG.
[0050] Figure 3 yes Figure 2 Flow chart of step S202 in FIG.
[0051] Figure 4 yes Figure 1 Flow chart of step S103 in FIG.
[0052] Figure 5 yes Figure 1 Flow chart of step S104 in FIG.
[0053] Figure 6 yes Figure 1 Flow chart of step S105 in FIG.
[0054] Figure 7 yes Figure 1 Flow chart of step S106 in FIG.
[0055] Figure 8It is a schematic diagram of the structure of a response device based on user objection provided in an embodiment of the present application;
[0056] Fig. 9 It is a schematic diagram of the hardware structure of the electronic device provided in the embodiment of the present application. DETAILED DESCRIPTION
[0057] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0058] It should be noted that, although the functional modules are divided in the device schematic diagram and the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first", "second", etc. in the specification, claims and the above drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.
[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.
[0060] First, some nouns involved in this application are analyzed:
[0061] Intent recognition is an important research direction in the field of natural language processing (NLP) and is a branch of artificial intelligence. Intent recognition focuses on the process of determining the user's intention from the user's language input. In practical applications, intent recognition is used to understand the user's specific needs or goals, and is commonly used in dialogue systems, voice assistants, and customer service automation. By analyzing the user's inquiries or instructions, the intent recognition system can extract key information, determine the user's demand category, and trigger corresponding actions or responses accordingly, thereby achieving effective interaction with human users.
[0062] User Profiling: It is an application based on data analysis and artificial intelligence technology to describe and classify user behaviors, preferences, needs and other characteristics. By analyzing the user's personal data, such as age, gender, occupation, purchase history, browsing behavior, etc., a detailed user feature model is constructed, and the user feature model is used for personalized recommendations, advertising targeting or customer service.
[0063] Emotion Keywords: refers to words used to represent and identify specific emotional states in text processing and sentiment analysis, such as happiness, sadness, anger, surprise, etc., which are used to help computer systems understand and classify emotional expressions in human language.
[0064] When customer service staff are rejected when selling insurance products to customers, it is difficult for them to quickly adjust their communication methods and service strategies to address their objections, which leads to a decrease in the recommendation rate of insurance products. Therefore, how to improve the success rate of insurance product recommendations has become an urgent problem to be solved.
[0065] Based on this, the embodiments of the present application provide a method and device for responding to user objections, an electronic device and a storage medium, aiming to improve the success rate of recommending insurance products.
[0066] The embodiments of the present application provide a method and device for responding to user objections, an electronic device, and a storage medium, which are specifically described through the following embodiments. First, the method for responding to user objections in the embodiments of the present application is described.
[0067] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. 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.
[0068] 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.
[0069] The answering method based on user objection provided in the embodiment of the present application relates to the field of text generation technology. The answering method based on user objection provided in the embodiment of the present application can be applied to a terminal, can be applied to a server side, or can be software running in a terminal or a server side. In some embodiments, the terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc.; the server side can be configured as an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can be an application that implements the answering method based on user objection, etc., but is not limited to the above forms.
[0070] The present application can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present application can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.
[0071] It should be noted that in each specific implementation of the present application, when it comes to the need to perform relevant processing based on data related to user identity or characteristics such as user information, user behavior data, user historical data, and user location information, the user's permission or consent will be obtained first, and the collection, use, and processing of these data will comply with relevant laws, regulations, and standards. In addition, when the embodiment of the present application needs to obtain the user's sensitive personal information, the user's separate permission or consent will be obtained through a pop-up window or by jumping to a confirmation page. After clearly obtaining the user's separate permission or consent, the necessary user-related data for the normal operation of the embodiment of the present application will be obtained.
[0072] Figure 1 is an optional flow chart of a method for responding to user objections provided in an embodiment of the present application. Figure 1 The method may include but is not limited to steps S101 to S106.
[0073] Step S101, obtaining target user information of a target object, obtaining target product information of a target product, and obtaining target objection information of a target object to a target product;
[0074] Step S102, constructing a target user portrait according to the target user information;
[0075] Step S103, performing intent recognition on the target objection information to obtain target intent information;
[0076] Step S104, performing emotion recognition on the target objection information to obtain target emotion information;
[0077] Step S105, generating an objection response using a preset objection response generation model, target user portrait, target intent information, and target product information to obtain an initial response text;
[0078] Step S106, respond based on the target emotion information and the initial response text.
[0079] Steps S101 to S106 shown in the embodiment of the present application obtain target user information of the target object, target product information and target objection information of the target product, then construct a target user portrait based on the target user information, and perform intent recognition on the target objection information. Then, a response is generated through a preset objection response generation model, the target user portrait, the target intent information and the target product information, thereby generating an initial response text with personalized responses to the product and the user, so as to quickly adjust the service strategy according to the customer's objection. At the same time, emotion recognition is performed on the target objection information to obtain target emotion information, and finally a response is performed through the target emotion information and the initial text, thereby improving the product recommendation success rate.
[0080] In step S101 of some embodiments, the target object is a customer who needs to purchase insurance in the insurance service, which may be an individual, a family or an enterprise, that is, a potential user of the insurance product. The target user information is the detailed information of the target object, including but not limited to the target object's age data, gender data, geographic location data, purchase history data, historical browsing data, occupation data and income data. The target product is an insurance product, and the target product information is the specific insurance terms of the insurance product. The target objection information refers to the questions, doubts or reasons for rejection raised by the target object regarding the target product.
[0081] See also Figure 2 In some embodiments, the target user information includes: age data, gender data, geographic location data, purchase history data, historical browsing data, occupation data and income data. Step S102 may include but is not limited to steps S201 to S204:
[0082] Step S201, constructing a user identity portrait based on age data, gender data and geographic location data;
[0083] Step S202, constructing a purchase tendency profile based on the purchase history data and the historical browsing data;
[0084] Step S203, constructing a purchasing power profile based on occupation data and income data;
[0085] Step S204, performing portrait aggregation based on the user identity portrait, purchase tendency portrait and purchasing power portrait to obtain a target user portrait.
[0086] Steps S201 to S204 shown in the embodiment of the present application construct a user identity portrait based on age data, gender data and geographic location data, then construct a purchase tendency portrait based on purchase history data and historical browsing data, and construct a purchasing power portrait based on occupation data and income data, and finally aggregate the user identity portrait, purchase tendency portrait and purchasing power portrait to obtain a target user portrait, so that customer service personnel have a deeper understanding and insight into customers when facing customer objections, so that they can quickly and effectively adjust communication strategies and service methods to match customers' specific needs and expectations, thereby improving the success rate of insurance product recommendations.
[0087] In step S201 of some embodiments, the user identity portrait is constructed through three dimensions. The first is to match the age data with the preset database to determine the probability of purchasing the target product in the age group in the historical data, and then determine the probability of purchasing the product in the historical data according to the gender data. Finally, the living environment is determined according to the geographic location, and the purchase probability determined by the age data and the purchase probability determined by the gender data and the living environment are used as the user identity portrait.
[0088] See also Figure 3 In some embodiments, step S202 may include but is not limited to steps S301 to S303:
[0089] Step S301, identifying product preferences based on purchase history data and historical browsing data to obtain product preference information;
[0090] Step S302, performing time series analysis based on the purchase history data and the historical browsing data to obtain product purchase trends;
[0091] Step S303: construct a purchase tendency profile based on product purchase trends and product preference information.
[0092] In steps S301 to S303 shown in the embodiment of the present application, product preference information is obtained by identifying product preferences based on purchase history data and historical browsing data, and then time series analysis is performed based on the purchase history data and historical browsing data to obtain product purchase trends. Finally, a purchase tendency portrait is constructed based on the product purchase trend and product preference information, thereby deeply understanding the potential purchase tendency of the target object and improving the recommendation success rate of insurance products.
[0093] In step S301 of some embodiments, the historical purchase data is the target object's historical purchase records, including but not limited to product type, purchase date, purchase frequency, purchase quantity, spending amount, etc. The historical browsing data is the target object's browsing activities on the website or application, such as the product pages viewed, dwell time, click path, search query, etc. Preference identification is performed based on the historical purchase data and historical browsing data, specifically, KNN clustering is performed based on the historical purchase data and historical browsing data, and then the product preference information of the target object for the product is determined based on the ratio between the data point in the cluster area and all the data points.
[0094] In step S302 of some embodiments, a time series analysis is performed based on the purchase history data and the historical browsing data, specifically, scheduling is performed based on the time of the purchase history data and the historical browsing data, the purchase history data and the historical browsing data are sorted by season, and then the browsing and purchasing frequency of the target object for the product in each season is determined, and the probability of the target object purchasing the product in each season is determined. Finally, based on the browsing and purchasing frequency of the target object and the current time, the purchase probability of the target object for the product, that is, the product purchase trend, is determined.
[0095] In step S303 of some embodiments, the product purchase trend and product preference information are merged to obtain a purchase tendency portrait of the target object for the target product.
[0096] In step S203 of some embodiments, the occupational data includes but is not limited to the target object's job title, industry type, job nature and occupational stability. The job title, industry type, job nature, occupational stability and income data are then input into a pre-trained support vector machine to derive the target object's purchase probability for the target product, i.e., the purchasing power portrait.
[0097] In step S204 of some embodiments, the user identity portrait, purchasing power tendency portrait and purchasing power portrait are concatenated to obtain a target user portrait.
[0098] See also Figure 4 In some embodiments, step S103 may include but is not limited to steps S401 to S403:
[0099] Step S401, performing semantic recognition on target objection information to obtain target objection semantics;
[0100] Step S402, classifying the target objection semantics according to the preset objection category to obtain the target objection category;
[0101] Step S403: perform intent recognition based on the target objection semantics and target objection category to obtain target intent information.
[0102] In steps S401 to S403 shown in the embodiment of the present application, semantic recognition is performed on the target objection information to obtain the target objection semantics, and then the target objection semantics are classified according to preset objection categories to obtain the target objection category. Finally, intent recognition is performed based on the target objection semantics and the target objection category to obtain the target intent information, thereby deeply understanding the type of feedback from the target object, and identifying the intention of the target object based on the type of feedback, thereby improving the success rate of recommending products.
[0103] In step S401 of some embodiments, semantic recognition is performed on the target objection information through a pre-trained neural network model to obtain the target objection semantics. In one embodiment, the pre-trained neural network model is a Transformer model, and the target objection information is input into the pre-trained Transformer model to obtain the target objection semantics.
[0104] In step S402 of some embodiments, keywords are extracted from the target objection semantics, and cosine similarity is calculated based on the keywords of the target objection semantics and preset objection categories, and the target objection category of the target objection semantics is determined based on the cosine similarity.
[0105] In step S403 of some embodiments, a corresponding pattern matching rule is pre-set for each target objection category, and the target objection semantics is identified according to the pattern matching rule to obtain the intent identification information. For example, the target objection information is: "I am worried that this insurance package will not cover all important medical services", the target objection information is classified as "product coverage doubts", and then the target objection information is pattern matched based on the pre-set regular expression of "product coverage doubts", that is, the pattern matching rule, to obtain "worry, not covered and important medical services", that is, the target intent information.
[0106] See also Figure 5 In some embodiments, step S104 includes but is not limited to steps S501 to S503:
[0107] Step S501, extracting keywords from target objection information to obtain target keyword information;
[0108] Step S502, identifying the emotional tendency of the target keyword information according to a preset emotional keyword dictionary to obtain the keyword emotional tendency;
[0109] Step S503: perform weighted calculation on the keyword sentiment tendency to obtain target sentiment information.
[0110] Steps S501 to S503 shown in the embodiment of the present application perform keyword extraction on the target objection information, then use a preset emotional keyword dictionary to identify the emotional tendency, and finally perform weighted calculation of the keyword emotional tendency to obtain the target emotional information, so as to accurately understand the emotions expressed by the target objection information, and adjust the communication strategy and service strategy based on the emotions, thereby improving the success rate of product recommendations.
[0111] In step S501 of some embodiments, keywords are extracted from the target objection information to obtain target keyword information. In one embodiment, keywords are extracted from the target objection information using a TF-IDF algorithm to obtain target keyword information.
[0112] In step S502 of some embodiments, the preset emotional keyword dictionary consists of multiple words and their corresponding weights, that is, a data pair of (word, emotion, weight). The target keyword information is filtered according to the preset emotional keyword dictionary, and the keywords in the target keywords that do not exist in the preset emotional keyword dictionary are deleted, and the weights corresponding to the deleted target keywords are obtained to obtain the keyword emotional tendency.
[0113] In step S503 of some embodiments, a weighted calculation is performed based on each emotion in the keyword emotion tendency and the corresponding weight to obtain the tendency value of each emotion, and the emotion corresponding to the maximum tendency value is obtained as the target emotion information.
[0114] See also Figure 6 In some embodiments, step S105 includes but is not limited to steps S601 to S604:
[0115] Step S601, performing text encoding on the target user portrait, target intent information and target product information through an encoding module to obtain text encoding information;
[0116] Step S602, extracting features from the text encoding information through an attention module to obtain text feature information;
[0117] Step S603, performing feature normalization on the text feature information through a normalization module to obtain text normalization information;
[0118] Step S604, performing residual connection on the text normalization information through the residual block to obtain the initial response text.
[0119] In steps S601 to S604 shown in the embodiment of the present application, all relevant information is first converted into a unified text encoding format through the encoding module, and then the encoded data is deeply extracted through the attention module. The generalization ability of the model is then enhanced through the processing of the normalization module. Finally, the final initial response text is obtained through the residual block, so as to generate a targeted preliminary response text and improve the recommendation success rate of insurance products.
[0120] In step S601 of some embodiments, the encoding module is a pre-trained Transformer encoder that has the ability to understand context and capture dependencies between elements in the text. The target user portrait, target intent information, and target product information are all input into the encoding module, which concatenates, encodes, and captures dependencies of all information to achieve deep semantic understanding and contextual relationship analysis of the overall input information, and outputs a matrix represented by numerical values, i.e., text encoding information.
[0121] In step S602 of some embodiments, the attention module is composed of a multi-head attention sub-module, which includes multiple attention sub-modules. The text encoding information is input into the multi-head attention sub-module, and the multiple attention sub-modules are processed in parallel. Finally, the outputs of all attention sub-modules are spliced, and the spliced vector is input into the linear layer to obtain the final output to obtain text feature information.
[0122] In step S603 of some embodiments, the normalization layer normalizes the text feature information to make the text feature information more generalizable and to stably generate text normalization information when facing a wide range of target user portraits, target intent information, and target product information.
[0123] In step S604 of some embodiments, the residual block is composed of multiple linear connection layers, and the multiple linear connection layers are jump-connected, so that the multiple linear connection layers can learn the potential information in the text normalization information, further improve the generalization ability, and generate high-quality initial response text when facing a wide range of customer objection information.
[0124] See also Figure 7 In some embodiments, step S106 may include but is not limited to steps S701 to S704:
[0125] Step S701, extracting emotional keywords from the initial response text to obtain initial emotional keywords;
[0126] Step S702, adjusting the vocabulary of the initial emotion keywords according to the target emotion information to obtain the target tone keywords;
[0127] Step S703, adjusting the initial response text based on the target tone keywords to obtain a target response text;
[0128] Step S704: respond based on the target response text.
[0129] Steps S701 to S704 shown in the embodiment of the present application extract emotional keywords from the initial response text, then adjust these keywords according to the target emotional information, and finally reconstruct the response text based on the adjusted tone keywords, and finally respond based on the target response text, thereby matching the emotion of the initial response text with the customer's emotion, thereby improving the recommendation probability of the insurance product.
[0130] In step S701 of some embodiments, keywords are extracted from the initial response text to obtain initial emotional keywords. In one embodiment, keywords are extracted from the initial response text using a TF-IDF algorithm to obtain initial emotional keywords.
[0131] In step S702 of some embodiments, the target emotional information is first converted through a preset emotional mapping table. For example, if the target emotional information is "anger", the corresponding converted emotion is "comfort" by searching the emotional mapping table, and then the vocabulary is adjusted according to the initial emotional keyword of "comfort" to obtain the target tone keyword.
[0132] In step S703 of some embodiments, the initial emotion keywords are replaced according to the target tone keywords to obtain the target response text.
[0133] In step S704 of some embodiments, a response is given to the target object based on the target response text. The target response text may be directly output to the target object, the target object may be responded to in voice form, or the customer service staff may respond to the target object based on the target response text as a reference.
[0134] See also Figure 8 The embodiment of the present application further provides a device for responding to user objections, which can implement the above-mentioned method for responding to user objections. The device includes:
[0135] An information acquisition module is used to acquire target user information of a target object, acquire target product information of a target product, and acquire target objection information of a target object to a target product;
[0136] A portrait building module is used to build a target user portrait based on target user information;
[0137] An intention recognition module is used to recognize the intention of the target objection information and obtain the target intention information;
[0138] The emotion recognition module is used to perform emotion recognition on the target objection information to obtain the target emotion information;
[0139] The response generation module is used to generate a response through a preset objection response generation model, target user portrait, target intent information and target product information to obtain an initial response text;
[0140] The text response module is used to respond based on the target emotional information and the initial response text.
[0141] The specific implementation of the device for responding to user objections is substantially the same as the specific implementation of the method for responding to user objections described above, and will not be described in detail herein.
[0142] The embodiment of the present application also provides an electronic device, the electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the above-mentioned method of responding to user objections when executing the computer program. The electronic device can be any intelligent terminal including a tablet computer, a car computer, etc.
[0143] See also Fig. 9 , Fig. 9 The hardware structure of an electronic device of another embodiment is illustrated, and the electronic device includes:
[0144] The processor 901 may be implemented by a general-purpose CPU (Central Processing Unit), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of the present application;
[0145] The memory 902 may be implemented in the form of a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 902 may store an operating system and other application programs. When the technical solution provided in the embodiments of this specification is implemented by software or firmware, the relevant program code is stored in the memory 902, and the processor 901 calls and executes the method for responding to user objections based on the embodiments of this application;
[0146] Input / output interface 903, used to implement information input and output;
[0147] Communication interface 904, used to realize communication interaction between the device and other devices, which can be realized through wired mode (such as USB, network cable, etc.) or wireless mode (such as mobile network, WI FI, Bluetooth, etc.);
[0148] A bus 905 that transmits information between various components of the device (e.g., the processor 901, the memory 902, the input / output interface 903, and the communication interface 904);
[0149] The processor 901 , the memory 902 , the input / output interface 903 and the communication interface 904 are connected to each other in communication within the device via a bus 905 .
[0150] An embodiment of the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the above-mentioned method for responding based on user objections is implemented.
[0151] The memory, as a non-transient computer-readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
[0152] The user objection-based response method, user objection-based response device, electronic device and storage medium provided in the embodiments of the present application obtain target user information of the target object, target product information and target objection information of the target product, then construct a target user portrait based on the target user information, and perform intent recognition on the target objection information, and then generate a response through a preset objection response generation model, target user portrait, target intent information and target product information, thereby generating an initial response text with personalized responses to products and users, and realizing rapid adjustment of service strategies based on customer objections. At the same time, emotion recognition is performed on the target objection information to obtain target emotion information, and finally a response is performed through the target emotion information and the initial text, thereby improving the product recommendation success rate.
[0153] The embodiments described in the embodiments of the present application are intended to more clearly illustrate the technical solutions of the embodiments of the present application and do not constitute a limitation on the technical solutions provided in the embodiments of the present application. Those skilled in the art will appreciate that with the evolution of technology and the emergence of new application scenarios, the technical solutions provided in the embodiments of the present application are also applicable to similar technical problems.
[0154] Those skilled in the art will appreciate that the technical solutions shown in the figures do not constitute a limitation on the embodiments of the present application, and may include more or fewer steps than shown in the figures, or a combination of certain steps, or different steps.
[0155] The device embodiments described above are merely illustrative, and the units described as separate components may or may not be physically separated, that is, they may be located in one place or distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0156] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices may be implemented as software, firmware, hardware, or a suitable combination thereof.
[0157] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.
[0158] It should be understood that in the present application, "at least one (item)" means one or more, and "plurality" means two or more. "And / or" is used to describe the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" can mean: only A exists, only B exists, and A and B exist at the same time, where A and B can be singular or plural. The character " / " generally indicates that the objects associated before and after are in an "or" relationship. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b or c can mean: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, c can be single or multiple.
[0159] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the above units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.
[0160] The units described above as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0161] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.
[0162] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including multiple instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, referred to as ROM), random access memory (Random Access Memory, referred to as RAM), disk or optical disk and other media that can store programs.
[0163] The preferred embodiments of the present invention are described above with reference to the accompanying drawings, but the scope of the rights of the present invention is not limited thereto. Any modification, equivalent substitution and improvement made by a person skilled in the art without departing from the scope and essence of the present invention should be within the scope of the rights of the present invention.
Claims
1. A method for responding to user objections, characterized in that: The method comprises: Acquire target user information of a target object, acquire target product information of a target product, and acquire target objection information of the target object to the target product; Building a target user portrait according to the target user information; Performing intent recognition on the target objection information to obtain target intent information; Performing emotion recognition on the target objection information to obtain target emotion information; Generate an objection response using a preset objection response generation model, the target user portrait, the target intent information, and the target product information to obtain an initial response text; Respond based on the target emotion information and the initial response text.
2. The method according to claim 1, characterized in that The target user information includes: age data, gender data, geographic location data, purchase history data, historical browsing data, occupation data and income data; the target user portrait is constructed based on the target user information, including: Constructing a user identity profile based on the age data, gender data and geographic location data; Building a purchase tendency profile based on the purchase history data and the historical browsing data; constructing a purchasing power profile based on the occupation data and the income data; The target user portrait is obtained by performing portrait aggregation based on the user identity portrait, the purchase tendency portrait and the purchasing power portrait.
3. The method according to claim 2, characterized in that The constructing a purchase tendency profile according to the purchase history data and the historical browsing data includes: Perform product preference identification based on the purchase history data and the historical browsing data to obtain product preference information; Performing time series analysis based on the purchase history data and the historical browsing data to obtain product purchase trends; The purchase tendency portrait is constructed according to the product purchase trend and the product preference information.
4. The method according to any one of claims 1 to 3, characterized in that: The responding based on the target emotion information and the initial response text includes: Extracting emotional keywords from the initial response text to obtain initial emotional keywords; Performing vocabulary adjustment on the initial emotion keywords according to the target emotion information to obtain target tone keywords; Performing text adjustment on the initial response text based on the target tone keywords to obtain a target response text; A response is made based on the target response text.
5. The method according to any one of claims 1 to 3, characterized in that: The performing intention recognition on the target objection information to obtain target intention information includes: Performing semantic recognition on the target objection information to obtain target objection semantics; Classifying the target objection semantics according to a preset objection category to obtain a target objection category; Intent recognition is performed based on the target objection semantics and the target objection category to obtain the target intent information.
6. The method according to any one of claims 1 to 3, characterized in that: The performing emotion recognition on the target objection information to obtain target emotion information includes: Performing keyword extraction on the target objection information to obtain target keyword information; According to a preset emotional keyword dictionary, the target keyword information is subjected to emotional tendency recognition to obtain the keyword emotional tendency; The keyword sentiment tendency is weightedly calculated to obtain the target sentiment information.
7. The method according to any one of claims 1 to 3, characterized in that: The objection response generation model includes: an encoding module, an attention module, a normalization module and a residual block; the objection response generation model is preset, the target user portrait, the target intent information and the target product information are used to generate the response, and the initial response text is obtained, including: Performing text encoding on the target user portrait, the target intent information and the target product information through the encoding module to obtain text encoding information; Extracting features of the text encoding information through the attention module to obtain text feature information; Performing feature normalization on the text feature information by the normalization module to obtain text normalization information; The text normalization information is subjected to residual connection through the residual block to obtain the initial response text.
8. A response device based on user objection, characterized in that: The device comprises: An information acquisition module is used to acquire target user information of a target object, acquire target product information of a target product, and acquire target objection information of the target object to the target product; A portrait building module, used to build a target user portrait according to the target user information; An intention recognition module, used to perform intention recognition on the target objection information to obtain target intention information; An emotion recognition module, used to perform emotion recognition on the target objection information to obtain target emotion information; A response generation module, used to generate a response through a preset objection response generation model, the target user portrait, the target intent information and the target product information to obtain an initial response text; A text response module is used to respond based on the target emotional information and the initial response text.
9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and the processor implements the method for responding to user objections according to any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for responding to user objections according to any one of claims 1 to 7 is implemented.