Customer service information determination method and device based on artificial intelligence technology

Abstract

The invention discloses a customer service information determination method and device based on an artificial intelligence technology, and the method comprises the steps: receiving multi-mode input information, and determining the type of the input information; according to the type of the input information, performing structured processing on the input information to determine structured information; extracting semantic information in the structured information by using a pre-established semantic analysis network; querying a dynamic knowledge base according to the semantic information to determine customer service response information; a new customer service solution is established based on the artificial intelligence technology, the understanding ability of the customer service robot can be improved, the question and answer efficiency is improved, complex machine learning training is avoided, operation resources are saved, and the customer experience is improved.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and device for determining customer service information based on artificial intelligence technology. Background Art

[0002] Currently, customer service has largely transitioned from traditional manual methods to a more human-machine approach. This involves using a customer service robot capable of natural language conversation to answer user questions. However, many customers are currently dissatisfied with the capabilities of these robots.

[0003] Currently, customer service robots are primarily based on the following technologies. The first is based on rule engines (such as keyword matching and decision trees), using predefined rules to handle simple question-and-answer inquiries. This type of customer service robot lacks the ability to handle complex scenarios and contextual understanding, resulting in very limited functionality. The second type primarily relies on static knowledge bases, searching for answers to various questions from the knowledge base to provide answers. However, this type of customer service robot is also rigid and may still be unable to effectively handle difficult questions, or may have difficulty understanding the semantics of difficult questions. Furthermore, the information in the knowledge base requires manual updating, which can lead to timeliness issues. The third type of customer service robot is based on artificial intelligence networks. These artificial intelligence networks require data training using machine learning algorithms to achieve their functionality. However, typical machine learning training requires a very specific focus, meaning that they can only effectively analyze specific areas in which they have been trained and are unable to analyze untrained content. This means that these networks lack versatility, and the complex training process consumes a large amount of historical data and computing resources.

[0004] It can be seen that in the existing technology, there is a lack of a powerful and versatile "customer service robot" solution. Summary of the Invention

[0005] The present invention provides a method and device for determining customer service information based on artificial intelligence technology, so as to provide a highly intelligent customer service solution.

[0006] In a first aspect, the present invention provides a method for determining customer service information based on artificial intelligence technology, comprising:

[0007] receiving multimodal input information and determining a type of the input information;

[0008] Performing structural processing on the input information according to the type of the input information to determine structured information;

[0009] Extracting semantic information from the structured information using a pre-established semantic analysis network;

[0010] The dynamic knowledge base is queried based on the semantic information to determine customer service response information.

[0011] Preferably, the type of the input information includes text information, voice information and image information, and the structuring of the input information according to the type of the input information to determine the structured information includes:

[0012] When the input information is voice information, performing voice recognition on the voice information to convert it into text information;

[0013] When the input information is image information, performing optical character recognition on the image information to convert it into text information;

[0014] The text information is cleaned and structured to determine the structured information.

[0015] Preferably, it also includes:

[0016] Establishing the semantic analysis network based on the pre-trained language model;

[0017] The semantic analysis network includes a short text analysis structure and a context analysis structure.

[0018] Preferably, the extracting semantic information from the structured information using a pre-established semantic analysis network includes:

[0019] Utilizing the short text analysis structure, extracting key features from the structured information;

[0020] Using the context analysis structure, determining context features based on historical conversations and user profiles;

[0021] The semantic information is determined according to the key features and the context features.

[0022] Preferably, it also includes:

[0023] The dynamic knowledge base is established based on multi-source knowledge information, target information, external environment data and user feedback data.

[0024] Preferably, it also includes:

[0025] The dynamic knowledge base is updated based on an event-driven strategy.

[0026] Preferably, querying the dynamic knowledge base according to the semantic information to determine the customer service response information includes:

[0027] querying the dynamic knowledge base according to the semantic information to determine the best response solution;

[0028] Generate natural language-based customer service response information based on the optimal response plan.

[0029] In a second aspect, the present invention provides a customer service information determination device based on artificial intelligence technology, comprising:

[0030] An information receiving module, configured to receive multimodal input information and determine the type of the input information;

[0031] a structuring module, configured to perform structural processing on the input information according to the type of the input information to determine structured information;

[0032] A semantic analysis module, configured to extract semantic information from the structured information using a pre-established semantic analysis network;

[0033] The customer service response module is used to query the dynamic knowledge base according to the semantic information to determine the customer service response information.

[0034] In a third aspect, the present invention provides a readable medium comprising an execution instruction. When a processor of an electronic device executes the execution instruction, the electronic device executes any method described in the first aspect.

[0035] In a fourth aspect, the present invention provides an electronic device comprising a processor and a memory storing execution instructions, wherein when the processor executes the execution instructions stored in the memory, the processor executes any method described in the first aspect.

[0036] The present invention provides a method and device for determining customer service information based on artificial intelligence technology, and establishes a new customer service solution based on artificial intelligence technology, which can improve the comprehension ability of customer service robots, increase question-and-answer efficiency, avoid complex machine learning training, save computing resources, and improve customer experience.

[0037] The further effects of the above-mentioned non-conventional preferred embodiment will be described below in conjunction with specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the existing technical solutions, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0039] Figure 1A flowchart of a method for determining customer service information based on artificial intelligence technology provided by one embodiment of the present invention;

[0040] Figure 2 A flowchart of another method for determining customer service information based on artificial intelligence technology provided by one embodiment of the present invention;

[0041] Figure 3 A schematic diagram of the structure of a customer service information determination device based on artificial intelligence technology provided by one embodiment of the present invention;

[0042] Figure 4 A schematic structural diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0044] Currently, customer service has largely transitioned from traditional manual methods to a more human-machine approach. This involves using a customer service robot capable of natural language conversation to answer user questions. However, many customers are currently dissatisfied with the capabilities of these robots.

[0045] Currently, customer service robots are primarily based on the following technologies. The first is based on rule engines (such as keyword matching and decision trees), using predefined rules to handle simple question-and-answer inquiries. This type of customer service robot lacks the ability to handle complex scenarios and contextual understanding, resulting in very limited functionality. The second type primarily relies on static knowledge bases, searching for answers to various questions from the knowledge base to provide answers. However, this type of customer service robot is also rigid and may still be unable to effectively handle difficult questions, or may have difficulty understanding the semantics of difficult questions. Furthermore, the information in the knowledge base requires manual updating, which can lead to timeliness issues. The third type of customer service robot is based on artificial intelligence networks. These artificial intelligence networks require data training using machine learning algorithms to achieve their functionality. However, typical machine learning training requires a very specific focus, meaning that they can only effectively analyze specific areas in which they have been trained and are unable to analyze untrained content. This means that these networks lack versatility, and the complex training process consumes a large amount of historical data and computing resources.

[0046] It can be seen that in the existing technology, there is a lack of a powerful and versatile "customer service robot" solution.

[0047] In view of this, the present invention provides a method for determining customer service information based on artificial intelligence technology. Figure 1 FIG. 1 is a specific embodiment of the method for determining customer service information based on artificial intelligence technology provided by the present invention. In this embodiment, the method includes:

[0048] Step 101: Receive multimodal input information and determine the type of the input information.

[0049] It can be considered that the execution subject of the method in this embodiment is a customer service system established based on artificial intelligence technology, in other words, a "customer service robot". Its main function is to receive "inquiries" sent by users and then generate corresponding "answers". The method in this embodiment supports "multimodal" information input, that is, users can input multiple types of information as "input information". For different types of input information, their specific types must first be identified before subsequent processing can be carried out. In this embodiment, the types of input information include text information, voice information, and image information.

[0050] Step 102: Perform structural processing on the input information according to the type of the input information to determine structured information.

[0051] In the customer service field, both inquiries and responses are often based on text. Related analysis and processing are often based on technologies such as text analysis and natural language analysis. Therefore, the subsequent processing of input information also centers around text processing. Non-textual input, such as voice and image information, requires conversion to text.

[0052] That is, when the input information is voice information, voice recognition is performed on the voice information to convert it into text information; when the input information is image information, optical character recognition is performed on the image information to convert it into text information. The conversion of voice information to text information can be achieved based on ASR (automatic speech recognition) technology. The conversion of image information to text information can be based on OCR (optical character recognition) technology. For input information that is itself of text type, no conversion is required and it can be processed directly. The text information obtained after the conversion is no different from the text information obtained directly in subsequent processing, and the same method can be used to complete structured processing. That is, the text information is subjected to data cleaning and structured processing to determine the structured information.

[0053] Structural processing, also known as normalization, ensures that text information conforms to specific data structures and formats (such as timestamp uniformity and encoding conversion) to facilitate more efficient subsequent analysis and computation. Before structural processing, data cleaning can also be performed to remove noise and improve precision and accuracy, ensuring that input data meets subsequent processing requirements.

[0054] In this step, you can use open-source toolkits (such as TensorFlow ASR and Tesseract OCR) to implement multimodal data parsing. This can be combined with lightweight edge computing devices (such as NVIDIA Jetson) to complete structured processing.

[0055] Step 103: Extract semantic information from the structured information using a pre-established semantic analysis network.

[0056] In this embodiment, a semantic analysis network will be pre-established based on artificial intelligence. The semantic analysis network can perform semantic analysis on the above-mentioned structured information and extract semantic information from it as the basis for the subsequent customer service robot to answer questions. In other words, it is necessary to establish the semantic analysis network based on a pre-trained language model. The pre-trained language model can specifically be a BERT network (i.e., Bidirectional Encoder Representation from Transformers) or GPT-3. Logically speaking, the semantic analysis network can adopt a hybrid architecture, mainly including two components, a short text analysis structure and a context analysis structure. The short text analysis structure is used to directly analyze structured information, determine the user's intentions (such as complaints, consultations, after-sales requests), and specific demands. The context analysis structure can realize context coherence analysis based on the context of the conversation and the user portrait, and dig out potential intentions that are not directly contained in the structured information, or further verify and verify the content in the structured information, so as to obtain more accurate and true semantic information.

[0057] For example, in some cases, a user first asks "What events are there in March?" and then asks "What about April?" Without considering the context, it is difficult to identify the user's true meaning by simply looking at the latter question. However, if we perform correlation analysis with the first question, we can infer that the true meaning of the latter question is "What events are there in April?"

[0058] From a technical perspective, short text analysis can use BERT for short text intent classification. Context analysis can use LSTM + Attention to process long text context. Knowledge graphs can also be integrated to assist in reasoning, enhancing understanding of professional terminology and business rules.

[0059] As can be seen from the above, the semantic analysis network in this embodiment is based on a large language model currently in use. Compared to existing technologies, it offers more comprehensive functionality and enhanced understanding capabilities. Furthermore, this foundation avoids the complex machine learning techniques used in existing technologies, improving efficiency and conserving computing resources.

[0060] The specific process of extracting semantic information from the structured information using the semantic analysis network is as follows: using the short text analysis structure to extract key features from the structured information; using the context analysis structure to determine context features based on historical conversations and user portraits; and determining the semantic information based on the key features and the context features.

[0061] Step 104: query the dynamic knowledge base based on the semantic information to determine customer service response information.

[0062] After determining the semantic information, a query can be performed based on the dynamic knowledge base, and the customer service response information (that is, the content of the customer service answer) can be determined based on the user's semantics. For example, when a user raises a "question", the semantic information represents the specific meaning of the "question". After knowing this specific meaning, the "answer" can be queried from the dynamic knowledge base. Then, this "answer" is provided to the user, completing the answering process. At this point, the method in this embodiment completes the response to the user's customer service information.

[0063] Through the above technical solutions, it can be seen that the beneficial effects of this embodiment are: establishing a new customer service solution based on artificial intelligence technology can improve the understanding ability of the customer service robot, increase the efficiency of question and answer, and avoid complex machine learning training, save computing resources, and improve customer experience.

[0064] Figure 1 What is shown is only a basic embodiment of the method of the present invention. By performing certain optimization and expansion on this basis, other preferred embodiments of the method can be obtained.

[0065] like Figure 2 FIG. 1 shows another specific embodiment of the method for determining customer service information based on artificial intelligence technology according to the present invention. This embodiment further describes the above embodiment and details the relevant features of the dynamic knowledge base. In this embodiment, the method includes the following steps:

[0066] Step 201: Receive multimodal input information and determine the type of the input information.

[0067] It can be considered that the execution subject of the method in this embodiment is a customer service system established based on artificial intelligence technology, in other words, a "customer service robot". Its main function is to receive "inquiries" sent by users and then generate corresponding "answers". The method in this embodiment supports "multimodal" information input, that is, users can input multiple types of information as "input information". For different types of input information, their specific types must first be identified before subsequent processing can be carried out. In this embodiment, the types of input information include text information, voice information, and image information.

[0068] Step 202: Perform structural processing on the input information according to the type of the input information to determine structured information.

[0069] In the customer service field, both inquiries and responses are often based on text. Related analysis and processing are often based on technologies such as text analysis and natural language analysis. Therefore, the subsequent processing of input information also centers around text processing. Non-textual input, such as voice and image information, requires conversion to text.

[0070] That is, when the input information is voice information, voice recognition is performed on the voice information to convert it into text information; when the input information is image information, optical character recognition is performed on the image information to convert it into text information. The conversion of voice information to text information can be achieved based on ASR (automatic speech recognition) technology. The conversion of image information to text information can be based on OCR (optical character recognition) technology. For input information that is itself of text type, no conversion is required and it can be processed directly. The text information obtained after the conversion is no different from the text information obtained directly in subsequent processing, and the same method can be used to complete structured processing. That is, the text information is subjected to data cleaning and structured processing to determine the structured information.

[0071] Structural processing, also known as normalization, ensures that text information conforms to specific data structures and formats (such as timestamp uniformity and encoding conversion) to facilitate more efficient subsequent analysis and computation. Before structural processing, data cleaning can also be performed to remove noise and improve precision and accuracy, ensuring that input data meets subsequent processing requirements.

[0072] In this step, you can use open-source toolkits (such as TensorFlow ASR and Tesseract OCR) to implement multimodal data parsing. This can be combined with lightweight edge computing devices (such as NVIDIA Jetson) to complete structured processing.

[0073] Step 203: Extract semantic information from the structured information using a pre-established semantic analysis network.

[0074] In this embodiment, a semantic analysis network will be pre-established based on artificial intelligence. The semantic analysis network can perform semantic analysis on the above-mentioned structured information and extract semantic information from it as the basis for the subsequent customer service robot to answer questions. In other words, it is necessary to establish the semantic analysis network based on a pre-trained language model. The pre-trained language model can specifically be a BERT network (i.e., Bidirectional Encoder Representation from Transformers) or GPT-3. Logically speaking, the semantic analysis network can adopt a hybrid architecture, mainly including two components, a short text analysis structure and a context analysis structure. The short text analysis structure is used to directly analyze structured information, determine the user's intentions (such as complaints, consultations, after-sales requests), and specific demands. The context analysis structure can realize context coherence analysis based on the context of the conversation and the user portrait, and dig out potential intentions that are not directly contained in the structured information, or further verify and verify the content in the structured information, so as to obtain more accurate and true semantic information.

[0075] For example, in some cases, a user first asks "What events are there in March?" and then asks "What about April?" Without considering the context, it is difficult to identify the user's true meaning by simply looking at the latter question. However, if we perform correlation analysis with the first question, we can infer that the true meaning of the latter question is "What events are there in April?"

[0076] From a technical perspective, short text analysis can use BERT for short text intent classification. Context analysis can use LSTM + Attention to process long text context. Knowledge graphs can also be integrated to assist in reasoning, enhancing understanding of professional terminology and business rules.

[0077] As can be seen from the above, the semantic analysis network in this embodiment is based on a large language model currently in use. Compared to existing technologies, it offers more comprehensive functionality and enhanced understanding capabilities. Furthermore, this foundation avoids the complex machine learning techniques used in existing technologies, improving efficiency and conserving computing resources.

[0078] The specific process of extracting semantic information from the structured information using the semantic analysis network is as follows: using the short text analysis structure to extract key features from the structured information; using the context analysis structure to determine context features based on historical conversations and user portraits; and determining the semantic information based on the key features and the context features.

[0079] Step 204: Establish the dynamic knowledge base based on the multi-source knowledge information, target information, external environment data and user feedback data.

[0080] In the method described in this embodiment, a dynamic knowledge base will be established in advance. When the customer service response information (that is, the content of the customer service answer) is determined based on the user's semantics. For example, when a user raises a "question", the semantic information represents the specific meaning of the "question". After knowing this specific meaning, you can query the dynamic knowledge base for its "answer". What is different from the existing technology is that the existing technology mostly uses a static knowledge base. The information in the static knowledge base needs to be manually updated, there is a risk of delay and omission, which may lead to timeliness issues.

[0081] The dynamic knowledge base in this embodiment is different from this. The dynamic knowledge base is established by combining multi-source knowledge information, targeted information (such as product manuals, policy documents), external environmental data (such as weather information, logistics information, etc.) and user feedback data. It can be seen that in the process of establishing the dynamic knowledge base, internal / external information and even the actual feedback from users are fully taken into consideration, which can better understand and respond to users' "inquiries" and give answers that are more in line with their needs. In the dynamic knowledge base, Elasticsearch can be used to achieve efficient full-text retrieval, combined with the Neo4j graph database to manage entity relationships.

[0082] Step 205: Update the data of the dynamic knowledge base based on the event-driven strategy.

[0083] In this embodiment, to ensure the timeliness of the dynamic knowledge base and its true "dynamic" nature, the knowledge base can also be updated in real time based on an event-driven mechanism (e.g., policy change triggering event-driven). Specifically, a message queue (such as Kafka) can be used to monitor business system change events to trigger incremental updates to the knowledge base. In this way, the dynamic knowledge base in this embodiment solves the drawbacks of manual data updates.

[0084] Step 206: query the dynamic knowledge base according to the semantic information to determine the best response solution.

[0085] Step 207: Generate natural language-based customer service response information according to the optimal response solution.

[0086] In this embodiment, the best solution is matched from the dynamic knowledge base based on semantic information and a natural language response is generated. It can also support multi-round dialogue management and dynamically adjust the response strategy (such as adding details when the user asks a follow-up question). Reinforcement learning (RL) can be further introduced to optimize the decision logic and iterate the model based on user satisfaction feedback. Specifically, a template engine (Jinja2) can be combined with a generative model (such as T5) to balance response accuracy and flexibility. Deploy an A / B testing framework to compare the conversion rates and user ratings of different strategies and continuously optimize model parameters.

[0087] like Figure 3 The figure shows a specific embodiment of the customer service information determination device based on artificial intelligence technology of the present invention. The device described in this embodiment is used to perform Figures 1-2 The physical device of the method. Its technical solution is essentially consistent with the above embodiment, and the corresponding description in the above embodiment is also applicable to this embodiment. The device in this embodiment includes:

[0088] The information receiving module 301 is configured to receive multimodal input information and determine the type of the input information.

[0089] The structuring module 302 is configured to perform structuring processing on the input information according to the type of the input information to determine structured information.

[0090] The semantic analysis module 303 is configured to extract semantic information from the structured information using a pre-established semantic analysis network.

[0091] The customer service response module 304 is configured to query the dynamic knowledge base according to the semantic information to determine customer service response information.

[0092] In addition Figure 3 Based on the embodiment shown, preferably, the present invention further includes:

[0093] The types of input information include text information, voice information and image information. The structuring module 302 includes:

[0094] The voice conversion unit 321 is configured to perform voice recognition on the voice information when the input information is voice information, thereby converting the voice information into text information.

[0095] The image conversion unit 322 is used to perform optical character recognition on the image information when the input information is image information, thereby converting the image information into text information.

[0096] The structuring unit 323 is configured to perform data cleaning and structuring processing on the text information to determine the structured information.

[0097] The semantic analysis module 303 includes:

[0098] The key feature extraction unit 331 is configured to extract key features from the structured information by utilizing the short text analysis structure.

[0099] The context feature determination unit 332 is configured to utilize the context analysis structure to determine context features based on historical conversations and user profiles.

[0100] The semantic information determining unit 333 is configured to determine the semantic information according to the key feature and the context feature.

[0101] Also includes:

[0102] The knowledge base establishment module 305 is used to establish the dynamic knowledge base according to multi-source knowledge information, target information, external environment data and user feedback data, and update the data of the dynamic knowledge base based on an event-driven strategy.

[0103] The customer service response module 304 includes:

[0104] A solution generating unit 341 is configured to query the dynamic knowledge base according to the semantic information to determine the best response solution;

[0105] The solution screening unit 342 is configured to generate natural language-based customer service response information according to the optimal response solution.

[0106] Figure 4 : This is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. Among them, the memory may include internal memory, such as high-speed random access memory (RAM), and may also include non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0107] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 4Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or one type of bus.

[0108] Memory is used to store execution instructions. Specifically, execution instructions are computer programs that can be executed. Memory can include internal memory and non-volatile memory, and provides execution instructions and data to the processor.

[0109] In one possible implementation, a processor reads corresponding execution instructions from a non-volatile memory into a memory and then executes them. Alternatively, the processor can obtain corresponding execution instructions from another device to logically form an artificial intelligence-based customer service information determination device. The processor executes the execution instructions stored in the memory to implement the artificial intelligence-based customer service information determination method provided in any embodiment of the present invention.

[0110] The present invention Figure 3 The method performed by the customer service information determination device based on artificial intelligence technology provided in the illustrated embodiment can be applied to a processor or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above method can be completed by an integrated logic circuit of hardware in the processor or instructions in the form of software. The above processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The various methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0111] The steps of the method disclosed in conjunction with the embodiments of the present invention can be directly implemented and executed by a hardware decoding processor, or by a combination of hardware and software modules within the decoding processor. The software modules can be located in a storage medium well-known in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. The storage medium is located in the memory, and the processor reads the information in the memory and, in conjunction with its hardware, completes the steps of the method described above.

[0112] The embodiment of the present invention further proposes a readable medium, which stores an execution instruction. When the stored execution instruction is executed by the processor of the electronic device, the electronic device can execute the customer service information determination method based on artificial intelligence technology provided in any embodiment of the present invention, and is specifically used to execute the following Figure 1 or Figure 2 The method shown.

[0113] The electronic device described in each of the aforementioned embodiments may be a computer.

[0114] Those skilled in the art will appreciate that the embodiments of the present invention may be provided as methods or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or a combination of software and hardware.

[0115] The various embodiments of the present invention are described in a progressive manner. Similar portions between the various embodiments can be referred to in conjunction with each other. Each embodiment focuses on the differences between the other embodiments. In particular, the device embodiments are generally similar to the method embodiments, so their description is relatively simple. For relevant portions, refer to the description of the method embodiments.

[0116] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.

[0117] The foregoing is merely an embodiment of the present invention and is not intended to limit the present invention. It will be apparent to those skilled in the art that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention are intended to be included within the scope of the claims of the present invention.

Claims

1. A customer service information determination method based on artificial intelligence technology, characterized in that: include: receiving multimodal input information and determining a type of the input information; Performing structured processing on the input information according to the type of the input information to determine structured information; Extracting semantic information from the structured information using a pre-established semantic analysis network; The dynamic knowledge base is queried based on the semantic information to determine customer service response information.

2. The method according to claim 1, characterized in that The type of the input information includes text information, voice information and image information, and the input information is structured according to the type of the input information to determine that the structured information includes: When the input information is voice information, performing voice recognition on the voice information to convert it into text information; When the input information is image information, performing optical character recognition on the image information to convert it into text information; The text information is cleaned and structured to determine the structured information.

3. The method according to claim 1, characterized in that Also includes: Establishing the semantic analysis network based on the pre-trained language model; The semantic analysis network includes a short text analysis structure and a context analysis structure.

4. The method according to claim 3, characterized in that The extracting semantic information from the structured information using a pre-established semantic analysis network includes: Utilizing the short text analysis structure, extracting key features from the structured information; Using the context analysis structure, determining context features based on historical conversations and user profiles; The semantic information is determined according to the key features and the context features.

5. The method according to claim 1, characterized in that: Also includes: The dynamic knowledge base is established based on multi-source knowledge information, target information, external environment data and user feedback data.

6. The method according to claim 5, characterized in that Also includes: The dynamic knowledge base is updated based on an event-driven strategy.

7. The method according to any one of claims 1 to 6, characterized in that: The querying of the dynamic knowledge base according to the semantic information to determine the customer service response information includes: querying the dynamic knowledge base according to the semantic information to determine the best response solution; Generate natural language-based customer service response information based on the optimal response plan.

8. A customer service information determination device based on artificial intelligence technology, characterized in that: include: An information receiving module, configured to receive multimodal input information and determine the type of the input information; a structuring module, configured to perform structural processing on the input information according to the type of the input information to determine structured information; A semantic analysis module, configured to extract semantic information from the structured information using a pre-established semantic analysis network; The customer service response module is used to query the dynamic knowledge base according to the semantic information to determine the customer service response information.

9. A computer-readable storage medium storing a computer program, wherein the computer program is used to execute the customer service information determination method based on artificial intelligence technology described in any one of claims 1 to 7.

10. An electronic device, comprising: processor; a memory for storing instructions executable by the processor; The processor is used to read the executable instructions from the memory and execute the instructions to implement the customer service information determination method based on artificial intelligence technology as described in any one of claims 1-7 above.

Images