CN116955562A - Intelligent customer service system based on artificial intelligence technology - Google Patents

Abstract

The invention relates to the technical field of data processing, in particular to an intelligent customer service system based on an artificial intelligence technology, which can automatically identify and solve customer problems by utilizing a deep learning technology and a natural language processing technology and provide accurate answers and solutions so as to improve customer satisfaction and business efficiency. The invention also comprises intelligent knowledge base management and automatic learning functions, so that the system can be improved continuously and adapt to the changing customer demands.

Description

Intelligent customer service system based on artificial intelligence technology

Technical Field

The invention relates to the technical field of data processing, in particular to an intelligent customer service system based on an artificial intelligence technology.

Background

Conventional customer support services often require manual operations, resulting in long response times, high costs, and error-prone. In addition, conventional systems have difficulty in handling a large number of customer questions and diversified language expressions. Failure to provide accurate answers and solutions results in low user satisfaction.

Traditional customer service systems may suffer from limitations such as inaccurate question understanding, inaccurate answers, or lack of personalization.

Current customer support services also exist that fail to recognize the emotional state of the customer and respond accordingly. The lack of a personalized interactive experience makes the customer unaware and careless. The lack of automatic learning functions cannot accommodate changes in customer demand.

Disclosure of Invention

The invention provides an intelligent customer service system based on an artificial intelligence technology, which aims to provide efficient and personalized user support services.

In order to achieve the purpose of the invention, the technical scheme adopted is as follows: the intelligent customer service system comprises a user input module, an automatic question and answer module, an intelligent knowledge base management module and an emotion recognition and emotion analysis module, wherein a problem or a request which is put forward by a user is input into the user input module through a text or voice form, the automatic question and answer module analyzes and processes the problem or the request by using a deep learning algorithm and a natural language processing module, the intelligent knowledge base management module retrieves related knowledge and information, the intelligent knowledge base management module comprises an intelligent knowledge base, the intelligent knowledge base is updated and expanded through two modes of manual operation and automatic learning, the automatic question and answer module and the intelligent knowledge base are used for matching and reasoning, the problem or the request of the user is responded, and simultaneously emotional response is added through the emotion recognition and emotion analysis module according to the emotion state which is expressed by the user.

As an optimization scheme of the invention, the natural language processing module comprises a lexical analysis unit, a syntactic analysis unit, a semantic understanding unit and a language generation unit, wherein the lexical analysis unit decomposes texts or voices into basic language units, and the basic language units comprise words, phrases and punctuations; the syntax analysis unit establishes a syntax structure tree of the sentence by analyzing the dependency relationship among the words in the sentence; the semantic understanding unit performs semantic analysis and understanding on sentences or texts; the language generation unit generates natural language text based on a given input and context.

The intelligent knowledge base comprises a knowledge representation unit, a knowledge storage structure unit, a knowledge retrieval unit, a knowledge updating and maintenance unit and a knowledge reasoning and recommending unit, wherein the knowledge representation unit is used for representing knowledge in the knowledge base, the knowledge storage structure unit is used for storing knowledge by using a data structure, the knowledge retrieval unit is used for rapidly retrieving related knowledge according to the query requirement of a user, the knowledge updating and maintenance unit is used for timely updating and maintaining the intelligent knowledge base, and the knowledge reasoning and recommending unit is used for deriving new knowledge or recommending related knowledge from the intelligent knowledge base according to the query and the requirement of the user.

As an optimization scheme of the invention, the knowledge representation unit comprises a rule representation, an ontology representation and a graph representation, wherein the rule representation represents knowledge by using logic rules, the ontology representation represents knowledge by using a semantic network or an ontology language, and the graph representation represents knowledge by using a graph structure.

As an optimization scheme of the invention, the automatic learning of the intelligent knowledge base comprises a data collection unit, a data preprocessing unit, a feature extraction unit, a model training unit, a model evaluation and optimization unit and an automatic learning real-time updating unit, wherein the data collection unit collects and acquires text data, and the text data comprises a history record of questions and answers, an intelligent knowledge base updating record and user feedback; the data preprocessing unit performs text cleaning, word segmentation or stop word removal operation on the collected text data; the feature extraction unit extracts useful features from the preprocessed text data to represent semantic information of questions and answers; the model training unit trains the extracted features by using a machine learning algorithm, and builds a question understanding and answer generating model; the model evaluation and optimization unit evaluates the trained question understanding and answer generation model, and verifies accuracy by comparing the model evaluation and optimization unit with a test data set; the automatic learning real-time updating unit updates the question understanding and answer generating model in real time.

As an optimization scheme, the emotion recognition and emotion analysis module comprises an emotion recognition unit, an emotion analysis unit and an emotion interaction unit, wherein the emotion recognition unit judges an emotion state by detecting keywords, modifier words and emotion words in sentences; the emotion analysis unit maps the emotion states to corresponding answer strategies; the emotional interaction unit flexibly uses language and expression to interact with the user.

As an optimization scheme of the invention, the emotion recognition and emotion analysis module further comprises a continuous learning and optimization unit and a safety and privacy protection unit, wherein the continuous learning and optimization unit improves the accuracy of question understanding and answer through interaction with an actual application scene and analysis of user feedback data, and also carries out automatic adjustment and optimization according to the user satisfaction degree and evaluation indexes of the question solving effect; the security and privacy protection unit uses encryption technology to protect the transmission and storage of user data, and anonymization and authority control are adopted to protect the privacy of personal information of the user.

The invention has the positive effects that: 1) The invention applies artificial intelligence technology to the field of customer support and provides efficient and personalized service. Through deep learning and natural language processing algorithms, the system is able to understand and answer complex customer questions and provide accurate solutions in a short time. The intelligent knowledge base management and automatic learning functions ensure that the system can continuously update and promote the knowledge base, and adapt to the continuously changing customer demands. The personalized service function enables the system to provide customized support according to the personalized requirements of clients, and user satisfaction and loyalty are improved. The emotion recognition and emotion analysis functions enable the system to better understand the emotion state of the client, and more humanized and emotional interaction experience is provided.

2) Compared with the traditional customer support mode, the intelligent customer service system has the following advantages:

high efficiency: the method can quickly identify and solve the problem of the client, realize instant response and high-efficiency processing, and reduce the waiting time of the client.

Accuracy: by means of deep learning and natural language processing algorithms, accurate answers and solutions can be provided, and errors and misleading possibly occurring in traditional manual operation are avoided.

Individualizing: personalized services can be provided according to the requirements and preferences of clients, and user experience and satisfaction are enhanced.

Learning ability: the knowledge base learning system has an automatic learning function, can continuously optimize and update the knowledge base, adapts to the change of the demands of clients, and improves the quality and efficiency of problem solving.

Emotional interaction: the emotion recognition and emotion analysis capability is provided, the emotion state of the client can be better understood, corresponding emotion interaction is provided, and emotion resonance and association of the user are enhanced.

Drawings

The invention will be described in further detail with reference to the drawings and the detailed description.

FIG. 1 is a schematic workflow diagram of the system of the present invention;

FIG. 2 is a flow chart of a system function implementation of the present invention.

Detailed Description

As shown in FIG. 1, the invention discloses an intelligent customer service system, which comprises a user input module, an automatic question-answering module, an intelligent knowledge base management module and an emotion recognition and emotion analysis module, wherein a question or a request which is put forward by a user is input into the user input module through a text or voice form, the automatic question-answering module analyzes and processes the question or the request by utilizing a deep learning algorithm and a natural language processing module, the intelligent knowledge base management module retrieves related knowledge and information, the intelligent knowledge base management module comprises an intelligent knowledge base, the intelligent knowledge base is updated and expanded through a manual operation mode and an automatic learning mode, the automatic question-answering module and the intelligent knowledge base are used for matching and reasoning to respond to the question or the request of the user, and simultaneously, the emotion response is increased through the emotion recognition and emotion analysis module according to the emotion state which is expressed by the user.

S1, user input: a user puts out a problem or a request to the intelligent customer service system through text, voice or other input modes; i.e. to communicate with the user via a network interface and to receive questions posed by the user.

S2, voice/text input processing: the input data is converted into a machine-understandable form by speech recognition or text parsing.

S3, intention recognition/voice understanding: the intelligent customer service system uses natural language processing technology to recognize and understand the intention input by the user and extract key information;

the questions are passed to an automatic question and answer module that analyzes and understands the questions using deep learning algorithms and natural language processing techniques. The intention and key information of the problem can be accurately identified.

S4, searching an intelligent knowledge base: the intelligent customer service system retrieves related knowledge and information from an intelligent knowledge base according to the user intention and the key information; the relevant knowledge and information is retrieved by the intelligent knowledge base management component. The intelligent knowledge base is composed of knowledge and data in multiple fields and can comprise common problem solutions, product manuals, technical documents and the like. The intelligent knowledge base can be updated and expanded by two modes of manual operation and automatic learning. Through the automatic learning function, the system can analyze and mine feedback data of the user, user behaviors and external information sources so as to maintain timeliness and accuracy of the knowledge base.

S5, generating a reply: based on the retrieved knowledge and information, the intelligent customer service system generates appropriate replies or suggestions in response to the user's questions and requests;

s6, replying and presenting: the generated reply is presented to the user by text, voice, or other suitable means;

s7, user feedback: the user provides further feedback and continues the conversation based on the replies provided by the intelligent customer service system.

When answering the questions, the system uses the automatic question answering module and the intelligent knowledge base to match and reason, and generates accurate answers or solutions. If the problem relates to personalized services, the system will provide personalized suggestions and recommendations based on the customer's history, preferences, and behavior patterns. In addition, according to the emotion and emotion expressed by the client, corresponding response can be made through emotion recognition and emotion analysis functions, so that interaction experience is enhanced.

S8, model optimization: the intelligent customer service system is used for continuously optimizing and improving by collecting and analyzing user feedback data, so that more accurate and effective reply is provided;

1. deep learning algorithm:

the intelligent customer service system of the invention utilizes a deep learning algorithm to analyze and understand the problems. Deep learning is a machine learning method based on an artificial neural network, and a complex data pattern is learned and represented through a multi-level neural network structure. The system uses a deep learning algorithm to encode and represent customer questions to obtain semantic information of the questions.

The basic principle of the deep learning algorithm is an artificial neural network. An artificial neural network is a mathematical model that simulates the structure and function of a biological nervous system, and is composed of a large number of artificial neurons (nodes) and connections between them.

Each artificial neuron receives inputs from a previous layer of neurons and weight sums the inputs through an activation function, and then passes the result to the next layer of neurons. Each connection has an associated weight for controlling the relative importance of the input signal. In addition, each neuron has a bias term for adjusting the response characteristics of the neuron.

The depth representation network (deep neural network) in the deep learning algorithm is made up of multiple levels, with the input layer receiving the raw data, the output layer producing the final predictions or results, and the intermediate hidden layer being responsible for extracting and combining the feature representations of the input data.

The core idea of deep learning is to enable the neural network to automatically learn and extract complex feature representations from a large amount of data by training its weight and bias parameters. This process is typically implemented by a gradient descent optimization algorithm that adjusts network parameters to minimize errors based on the errors between the network predicted results and the actual results.

In the training process, the deep learning algorithm calculates a prediction result through forward propagation according to the characteristics of input data and label information, and compares the prediction result with an actual label. The contribution of each weight and bias to the error is then calculated by a back propagation algorithm and their values are updated accordingly to reduce the error. This process is repeated until a predefined stopping condition is reached, such as a certain training round or error convergence.

By combining multi-layer networks and feature extraction layer by layer, deep learning algorithms are able to learn increasingly abstract and advanced feature representations. This hierarchical feature extraction capability enables deep learning to achieve significant success in processing complex data tasks such as images, speech, natural language, etc.

In terms of speech recognition, a speech recognition model can be trained to convert input sound into text by means of a deep learning algorithm. The model may be constructed by acoustic models and language models in deep learning techniques. In the training process, a large amount of voice data and text labels can be used, so that the recognition accuracy of the model for different accents and speech speeds is improved.

In the aspect of natural language processing, the operations such as grammar analysis, entity recognition, emotion analysis and the like can be performed on natural language problems through a deep learning algorithm. These operations may be performed by deep learning models such as convolutional neural networks, recurrent neural networks, or transformers. During the training process, a large amount of natural language data, such as text question-answering data, dialogue data and the like, can be used, so that the processing capacity of the model for different problems is improved.

In the aspect of constructing the knowledge graph, the knowledge graph can be constructed and trained by using models such as a graph neural network and the like through a deep learning algorithm. In the training process, a large amount of structured data and knowledge bases, such as business databases inside enterprises, knowledge bases on the public internet, and the like, can be used, so that the system can realize more accurate and comprehensive knowledge storage and reasoning capability.

In terms of problem classification and recognition, a classifier can be trained to classify and recognize problems posed by customers through a deep learning algorithm. The classifier can be constructed by a deep learning model such as a convolutional neural network or a cyclic neural network. In the training process, a large amount of labeling data can be used, for example, the problems of the clients are labeled according to the types of the problems, so that the classifier can accurately classify and identify the problems of different types.

In terms of semantic understanding, a semantic understanding model can be trained through a deep learning algorithm to convert the customer's problem into a computer-understandable language. The model can be constructed by a deep learning model such as a cyclic neural network or an attention mechanism. During the training process, a large amount of labeling data can be used, for example, the questions of the clients and the answers corresponding to the questions are labeled, so that the model can effectively answer the questions.

In the aspect of automatic answer generation, a generation model can be trained through a deep learning algorithm, and answers are automatically generated according to the questions and the context information of the clients. The model can be constructed by a deep learning model such as a cyclic neural network or a transducer. In the training process, a large amount of labeling data can be used, for example, the input questions and the corresponding answers thereof are labeled, so that the model can effectively answer the questions.

2. Natural language processing technology:

the system uses natural language processing technology to perform semantic matching and answer generation on the client questions. Natural language processing is a field of computer science that processes human language, covering tasks such as text analysis, semantic understanding, language generation, and the like. The system uses natural language processing techniques to analyze the semantics and intent of the customer's questions, match them to information in the knowledge base, and generate accurate answers or solutions.

The technology refers to the technical field of understanding, analyzing and processing natural language through a computer. It relates to a number of tasks and techniques including lexical analysis, syntactic analysis, semantic understanding, language generation, etc. The following are some basic principles of natural language processing techniques:

lexical Analysis (Lexical Analysis): lexical analysis is the decomposition of text into basic linguistic units, such as words, phrases, or punctuation marks. This process includes lexical division (token) and Part-of-Speech Tagging (Part-of-Speech Tagging). The lexical analyzer segments the input sentence into words according to grammatical rules and models of the language and assigns each word an appropriate part-of-speech tag, such as nouns, verbs, adjectives, etc.

Syntactic analysis (Syntactic Analysis): syntactic analysis is the process of studying the structure and grammatical relations of sentences. It builds a grammar tree of sentences by analyzing the dependency relationship between words in sentences. Common methods of syntactic analysis include rule-based syntactic analysis and statistical syntactic analysis. Syntactic analysis helps to understand the grammatical relations between words in a sentence and the structure of the sentence.

Semantic understanding (Semantic Understanding): semantic understanding refers to the process of semantic analysis and understanding of sentences or text. It involves tasks such as word sense disambiguation (Word Sense Disambiguation), named entity recognition (Named Entity Recognition), relationship extraction (Relation Extraction), etc. The goal of semantic understanding is to understand the meaning and context of sentences and extract key information from them.

Language generation (Language Generation): language generation is the process of generating natural language text according to certain rules or models. It involves the tasks of text synthesis, automatic summarization of text, machine translation, etc. The goal of language generation is to generate natural fluent text based on a given input and context.

These basic principles are the core components of natural language processing technology. By combining these principles and techniques, a natural language processing system is able to analyze, understand, and generate natural language, enabling automatic processing and semantic understanding of text. The techniques have wide application in the fields of machine translation, information retrieval, text classification, emotion analysis, intelligent dialogue, and the like.

3. Intelligent knowledge base management:

the knowledge base in the intelligent customer service system is a core component that stores and manages knowledge and information. The knowledge base contains knowledge and data in multiple fields, such as common problem solutions, product manuals, technical documents, and the like. The system stores the knowledge in a structured way through an intelligent knowledge base management function and provides an efficient retrieval and updating mechanism. The knowledge base can be manually operated by professionals, and can also be updated and expanded through an automatic learning function.

Intelligent knowledge base management refers to a management process of effectively organizing, storing, retrieving and updating a knowledge base to support the utilization and application of knowledge by an intelligent system. It relates to the basic principles of knowledge representation, storage structure, search algorithm, etc. The following are some basic principles of intelligent knowledge base management:

knowledge representation: knowledge in the knowledge base needs to be represented and stored in a suitable form in order for the system to understand and apply. Common knowledge representation methods include rule representation, ontology representation, and graph representation. The rule representation represents knowledge using logical rules, the ontology representation represents knowledge using semantic networks or ontology languages, and the graph representation represents knowledge using graph structures. Selection of an appropriate knowledge representation method can better support knowledge storage and reasoning.

Knowledge storage structure: intelligent knowledge bases typically use specific data structures to store knowledge. Common data structures include relational databases, graph databases, document databases, and the like. Selecting a suitable storage structure can improve knowledge retrieval efficiency and query performance.

And (5) knowledge retrieval: intelligent knowledge base management needs to provide an efficient retrieval mechanism so that the system can quickly retrieve relevant knowledge according to the query needs of the user. The retrieval method can be based on keyword matching, semantic matching, fuzzy matching and the like. Keyword matching is the most common search method, which matches in a knowledge base based on keywords entered by a user. The semantic matching is performed according to the semantic information of sentences or questions, so that the related knowledge can be acquired more accurately.

Knowledge updating and maintaining: the intelligent knowledge base needs to be updated and maintained regularly to ensure the timeliness and accuracy of the knowledge. The updating of knowledge can be performed by means of manual updating, automatic extraction, machine learning and the like. Maintenance includes verification, modification and deletion of knowledge to ensure quality and consistency of the knowledge base.

Knowledge reasoning and recommendation: the intelligent knowledge base management can also combine reasoning and recommendation technology to deduce new knowledge or recommendation related knowledge from the knowledge base according to the inquiry and the requirement of the user. Inference can be based on logic rules, inference mechanisms, or machine learning algorithms, and recommendation can be based on collaborative filtering, content filtering, or deep learning techniques.

4. Automatic learning function:

as shown in fig. 2, the system (intelligent customer service system) has an automatic learning function, and learns and updates a knowledge base from a large number of data sources through machine learning and data mining technologies. The system may analyze the user's feedback data, user behavior, and external information sources to discover new knowledge and information and automatically integrate it into the knowledge base. The automatic learning function enables the system to continuously optimize and update the knowledge base to adapt to the change of the customer demand.

In an intelligent customer service system, the basic principle of the automatic learning function comprises the following steps:

and (3) data collection: the system collects and obtains a large amount of text data including histories of questions and answers, knowledge base update records, user feedback, and the like. These data are used to train and update the model of the system.

Data preprocessing: and preprocessing the collected data, including text cleaning, word segmentation, stop word removal and other operations. The purpose of the preprocessing is to remove noise, normalize the text data for subsequent feature extraction and model training.

Feature extraction: useful features are extracted from the pre-processed text data to represent semantic information of questions and answers. Common feature extraction methods include Word bag models, TF-IDF vectors, word Embedding (Word Embedding), and the like.

Model training: the extracted features are trained using machine learning algorithms (e.g., neural networks, support vector machines, decision trees, etc.), and a question understanding and answer generation model is constructed. The training process is to adjust model parameters so that they can accurately classify questions and answer the questions.

Model evaluation and optimization: and evaluating the trained model, and verifying the accuracy and the performance of the model through the verification with the test data set. And optimizing and adjusting the model according to the evaluation result to improve the effect and generalization capability of the model.

Updating in real time: the system updates the model in real time according to the new data source and information. This may be done by periodic model training and batch updates, or by incremental learning and online learning.

5. Emotion recognition and emotion analysis:

the system has emotion recognition and emotion analysis functions, and can recognize emotion and emotion in customer expression. Emotion recognition utilizes natural language processing and machine learning techniques to determine emotional states in customer questions, such as happiness, disappointment, anger, etc. And the emotion analysis makes corresponding response according to the emotion state of the client so as to provide more emotional and personalized interactive experience.

Emotion recognition technology: the system uses emotion recognition techniques to recognize emotions and moods that the customer embodies when expressing the problem. The technology utilizes natural language processing and machine learning algorithms to analyze the characteristics of words, mood, intonation and the like in the customer problem so as to judge the emotion state of the customer. For example, by detecting keywords, modifiers, and emotion words in a sentence, the system can automatically identify the emotional tendency of the customer.

Emotion analysis technology: after the emotion state of the client is identified, the system responds correspondingly through emotion analysis technology. This technique maps the emotional state of the client to a corresponding answer strategy based on machine learning and natural language processing algorithms. For example, when a customer expresses dissatisfaction or anger, the system may provide a priority scheme for solving the problem, a response to pacifying, or timely transfer the customer to an artificial customer service according to a policy set in advance.

Emotional interaction design: the system designs an emotional interaction mechanism to enhance the user experience. For example, in answering questions, the system can flexibly employ language and expressions to interact with clients to better convey emotions and establish emotion connections. The system can also use an emotional language model to make the answers more humanized and affinity to meet the customer's expectations for emotional services.

Continuous learning and optimization: the system adopts a continuous learning and optimizing mechanism to continuously improve the performance and effect of the intelligent customer service system. Through interaction with the actual application scene and analysis of user feedback data, the system can update the model and algorithm in real time, and accuracy of question understanding and answer is improved. The system can also automatically adjust and optimize according to the evaluation indexes of customer satisfaction and problem solving effect so as to provide better service.

Security and privacy protection: the system pays attention to security and privacy protection, and a series of measures are taken to ensure the security and confidentiality of customer data. The system uses encryption technology to protect the transmission and storage process of client data, and strictly complies with related privacy policies and laws and regulations. Meanwhile, the system also adopts measures such as data anonymization and authority control to protect the privacy of the personal information of the client.

An automatic question-answering module: by using natural language processing techniques and deep learning algorithms, the system is able to understand questions posed by a customer and automatically retrieve and generate accurate answers. The module performs training and optimization based on a large-scale knowledge base and corpus to improve question understanding and answer quality.

2. Intelligent knowledge base management: has a dynamic knowledge base for storing and managing knowledge and information in various fields. Can be updated and expanded by two modes of manual operation and automatic learning so as to ensure the timeliness and accuracy of knowledge.

3. Personalized service: personalized services can be provided according to the specific needs and preferences of the customer. By analyzing the history, behavior patterns and feedback information of the client, the system can automatically recommend related solutions and products, and better user experience is provided.

4. Emotion recognition and emotion analysis: can recognize emotion and emotion in the customer expression and respond accordingly. This enables the system to better understand the needs and emotional state of the customer, providing a more humanized and emotional support service.

The intelligent customer service system based on the artificial intelligence technology has remarkable technical effects:

firstly, the method can realize efficient and accurate problem solving and greatly improve the response speed and the solving efficiency of customer support. This helps to reduce customer latency and enhance customer satisfaction and loyalty.

Secondly, the intelligent knowledge base management and automatic learning functions enable the system to learn and improve continuously, and adapt to the continuously changing customer requirements and industry development. The knowledge base can be automatically updated and expanded, so that the system is ensured to always have the latest knowledge and information, and accurate and comprehensive support service is provided.

In addition, the personalized services and emotional interaction functions enhance the interactive experience between the system and the client. The interactive relationship between the structure and the components of the intelligent customer service system is a unique innovation. Efficient question understanding and answering capabilities are achieved through deep learning and natural language processing techniques. Through intelligent knowledge base management and automatic learning functions, knowledge can be continuously learned and updated, and the method is suitable for continuously changing customer demands. The personalized service and emotion recognition function promote user experience, so that the system can better understand the requirements and emotion states of clients and provide customized support.

The intelligent customer service system has wide application prospect in the software industry. The method can be applied to various scenes such as online customer service platforms, mobile application programs, intelligent voice assistants and the like, and provides efficient and personalized user support services. In the electronic commerce industry, the system can help clients to quickly solve the problem, provide accurate product information and recommendation, and promote shopping experience and sales growth. In the financial industry, the system can provide timely account inquiry, transaction guidance and risk management support, and user trust and satisfaction are enhanced. In the telecommunications industry, the system can handle a large number of user queries and fault reports, provide rapid technical support and fault removal, and improve the quality of service. While the foregoing is directed to embodiments of the present invention, other and further details of the invention may be had by the present invention, it should be understood that the foregoing description is merely illustrative of the present invention and that no limitations are intended to the scope of the invention, except insofar as modifications, equivalents, improvements or modifications are within the spirit and principles of the invention.

Claims (7)

1. An intelligent customer service system based on an artificial intelligence technology is characterized in that: the intelligent customer service system comprises a user input module, an automatic question-answering module, an intelligent knowledge base management component and an emotion recognition and emotion analysis module, wherein a problem or a request which is put forward by a user is input into the user input module in a text or voice mode, the automatic question-answering module analyzes and processes the problem or the request by utilizing a deep learning algorithm and a natural language processing module, the intelligent knowledge base management component retrieves related knowledge and information, the intelligent knowledge base management component comprises an intelligent knowledge base, the intelligent knowledge base is updated and expanded in a manual operation and automatic learning mode, the problem or the request of the user is responded through matching and reasoning of the automatic question-answering module and the intelligent knowledge base, and simultaneously emotional response is increased through the emotion recognition and emotion analysis module according to the emotion state which is expressed by the user.

2. An intelligent customer service system based on artificial intelligence technology as claimed in claim 1, wherein: the natural language processing module comprises a lexical analysis unit, a syntactic analysis unit, a semantic understanding unit and a language generating unit, wherein the lexical analysis unit decomposes texts or voices into basic language units, and the basic language units comprise words, phrases and punctuations; the syntax analysis unit establishes a syntax structure tree of the sentence by analyzing the dependency relationship among the words in the sentence; the semantic understanding unit performs semantic analysis and understanding on sentences or texts; the language generation unit generates natural language text based on a given input and context.

3. An intelligent customer service system based on artificial intelligence technology as claimed in claim 2, wherein: the intelligent knowledge base comprises a knowledge representation unit, a knowledge storage structure unit, a knowledge retrieval unit, a knowledge updating and maintaining unit and a knowledge reasoning and recommending unit, wherein the knowledge representation unit is used for representing knowledge in the knowledge base, the knowledge storage structure unit uses a data structure to store knowledge, the knowledge retrieval unit is used for rapidly retrieving related knowledge according to the query requirement of a user, the knowledge updating and maintaining unit is used for timely updating and maintaining the intelligent knowledge base, and the knowledge reasoning and recommending unit is used for deriving new knowledge or recommending related knowledge from the intelligent knowledge base according to the query and requirement of the user.

4. An intelligent customer service system based on artificial intelligence technology as claimed in claim 3, wherein: the knowledge representation unit comprises a rule representation, an ontology representation and a graph representation, wherein the rule representation represents knowledge by using logic rules, the ontology representation represents knowledge by using a semantic network or an ontology language, and the graph representation represents knowledge by using a graph structure.

5. The intelligent customer service system based on artificial intelligence technology as claimed in claim 4, wherein: the automatic learning of the intelligent knowledge base comprises a data collecting unit, a data preprocessing unit, a feature extracting unit, a model training unit, a model evaluating and optimizing unit and an automatic learning real-time updating unit, wherein the data collecting unit collects and acquires text data, and the text data comprises a history record of questions and answers, an intelligent knowledge base updating record and user feedback; the data preprocessing unit performs text cleaning, word segmentation or stop word removal operation on the collected text data; the feature extraction unit extracts useful features from the preprocessed text data to represent semantic information of questions and answers; the model training unit trains the extracted features by using a machine learning algorithm, and builds a question understanding and answer generating model; the model evaluation and optimization unit evaluates the trained question understanding and answer generation model, and verifies accuracy by comparing the model evaluation and optimization unit with a test data set; the automatic learning real-time updating unit updates the question understanding and answer generating model in real time.

6. An intelligent customer service system based on artificial intelligence technology as claimed in claim 1, wherein: the emotion recognition and emotion analysis module comprises an emotion recognition unit, an emotion analysis unit and an emotion interaction unit, wherein the emotion recognition unit judges an emotion state by detecting keywords, modifier words and emotion words in sentences; the emotion analysis unit maps the emotion states to corresponding answer strategies; the emotional interaction unit flexibly uses language and expression to interact with the user.

7. The intelligent customer service system based on artificial intelligence technology as claimed in claim 6, wherein: the emotion recognition and emotion analysis module further comprises a continuous learning and optimization unit and a safety and privacy protection unit, wherein the continuous learning and optimization unit improves the accuracy of question understanding and answer through interaction with an actual application scene and analysis of user feedback data, and automatically adjusts and optimizes according to evaluation indexes of user satisfaction and a question solving effect; the security and privacy protection unit uses encryption technology to protect the transmission and storage of user data, and anonymization and authority control are adopted to protect the privacy of personal information of the user.