Expert system knowledge base construction method, question-answering method, system, device and medium

Abstract

The present invention discloses an expert system knowledge base construction method, question-and-answer method, system, device, and medium. The construction method includes: obtaining preset first question-and-answer information, the first question-and-answer information including multiple first question descriptions and question types and first answer contents corresponding to the first question descriptions; performing feature extraction on the first question descriptions to obtain a question feature vector; determining a training data set based on the question feature vector, the question type, and the first answer contents, and then inputting the training data set into a pre-constructed deep neural network for training to obtain a trained knowledge reasoning model; periodically obtaining expert knowledge data based on the knowledge reasoning model, and then constructing an expert system knowledge base based on the expert knowledge data. The present invention improves the construction and update efficiency of the expert system knowledge base, reduces the computing power requirements of the system, thereby reducing the cost of expert system construction and improving the efficiency of question-and-answer communication. The present invention can be widely applied to the field of artificial intelligence technology.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an expert system knowledge base construction, identification method, system, device and medium. Background Art

[0002] An expert system is an intelligent computer program system that contains a wealth of expert-level knowledge and experience in a specific field. It can leverage the knowledge and problem-solving methods of human experts to solve problems in that field. In other words, an expert system is a program system with a wealth of specialized knowledge and experience. It applies artificial intelligence and computer technology to reason and judge based on the knowledge and experience provided by one or more experts in a specific field, simulating the decision-making process of human experts in order to solve complex problems that require human expertise. In short, an expert system is a computer program system that simulates the way human experts solve problems in a specific field.

[0003] Expert systems typically consist of six components: a human-computer interface, a knowledge base, an inference engine, an interpreter, a comprehensive database, and knowledge acquisition. The separation of the knowledge base and the inference engine is particularly distinctive. The architecture of an expert system varies depending on its type, function, and scale.

[0004] The expert system knowledge base is one of the core elements of an expert system. Its primary function is to store and manage the knowledge within the expert system, primarily consisting of knowledge from textbooks and the experiential knowledge gained by experts in various fields through long-term work practice. Currently, the vast majority of expert system knowledge bases are obtained through communication between knowledge engineers and human experts. Knowledge engineers obtain knowledge from human experts and encode it into the expert system. Existing technologies have the following shortcomings: the knowledge base content system has a slow update cycle and cannot be updated in a timely manner by the industry; the inference engine requires enormous computing power and is computationally expensive; and the human-computer interaction interface relies on traditional text input for retrieval, resulting in low communication efficiency and unfavorable application and promotion on smart terminal devices. Summary of the Invention

[0005] The purpose of the present invention is to solve one of the technical problems existing in the prior art to at least a certain extent.

[0006] Therefore, an object of an embodiment of the present invention is to provide a method for constructing an expert system knowledge base, which can improve the construction efficiency and update efficiency of the expert system knowledge base.

[0007] Another object of the embodiment of the present invention is to provide an efficient expert system knowledge base question-answering method.

[0008] In order to achieve the above technical objectives, the technical solutions adopted by the embodiments of the present invention include:

[0009] In a first aspect, an embodiment of the present invention provides a method for constructing an expert system knowledge base, comprising the following steps:

[0010] Obtaining preset first question-and-answer information, where the first question-and-answer information includes a plurality of first question descriptions and question types and first answer contents corresponding to the first question descriptions;

[0011] Performing feature extraction on the first problem description to obtain a problem feature vector;

[0012] Determining a training data set based on the question feature vector, the question type, and the first answer content, and then inputting the training data set into a pre-built deep neural network for training to obtain a trained knowledge reasoning model;

[0013] Expert knowledge data is periodically acquired according to the knowledge reasoning model, and then an expert system knowledge base is constructed according to the expert knowledge data.

[0014] Furthermore, in one embodiment of the present invention, the step of extracting features from the first problem description to obtain a problem feature vector specifically includes:

[0015] Performing word segmentation on the first problem description to obtain a first phrase;

[0016] Determine a word embedding vector for each word in the first phrase using a Word2ve model, and then determine a phrase vector for the first phrase based on the word embedding vector;

[0017] The phrase vector is normalized to obtain the question feature vector.

[0018] Furthermore, in one embodiment of the present invention, the step of determining a training data set based on the question feature vector, the question type, and the first answer content specifically includes:

[0019] Determine a training sample according to the problem feature vector;

[0020] Determine a primary label based on the question type, and determine a secondary label based on the first answer content;

[0021] A training data set is determined according to the primary labels and secondary labels corresponding to the training samples.

[0022] Furthermore, in one embodiment of the present invention, the step of inputting the training data set into a pre-built deep neural network for training specifically includes:

[0023] Inputting the training data set into the deep neural network to obtain question type prediction results and answer content prediction results;

[0024] Determining a training loss value based on the question type prediction result, the answer content prediction result, the primary label, and the secondary label;

[0025] The parameters of the deep neural network are updated according to the loss value.

[0026] Furthermore, in one embodiment of the present invention, the step of periodically acquiring expert knowledge data according to the knowledge reasoning model and then constructing an expert system knowledge base according to the expert knowledge data specifically includes:

[0027] Acquire a third problem description at a preset period, and reason and classify the third problem description according to the knowledge reasoning model to obtain expert knowledge data and a knowledge type of the expert knowledge data;

[0028] An expert system knowledge base is constructed or updated according to the expert knowledge data and the knowledge type.

[0029] In a second aspect, an embodiment of the present invention provides an expert system knowledge base question-answering method, comprising the following steps:

[0030] Acquiring user voice information and performing voice recognition on the user voice information to obtain a second question description;

[0031] Performing a matching search on the second question description using the expert system knowledge base obtained by the expert system knowledge base construction method according to the first aspect to obtain a second answer content;

[0032] or,

[0033] The second question description is inferred using the knowledge reasoning model obtained by the expert system knowledge base construction method as described in the first aspect to obtain the second answer content.

[0034] In a third aspect, an embodiment of the present invention provides an expert system knowledge base construction system, including:

[0035] A question and answer information acquisition module is used to acquire preset first question and answer information, where the first question and answer information includes a plurality of first question descriptions and question types and first answer contents corresponding to the first question descriptions;

[0036] A feature extraction module, configured to extract features from the first problem description to obtain a problem feature vector;

[0037] a model training module, configured to determine a training data set based on the question feature vector, the question type, and the first answer content, and then input the training data set into a pre-built deep neural network for training to obtain a trained knowledge reasoning model;

[0038] The knowledge base construction module is used to periodically acquire expert knowledge data according to the knowledge reasoning model, and then construct an expert system knowledge base according to the expert knowledge data.

[0039] In a fourth aspect, an embodiment of the present invention provides an expert system knowledge base question-answering system, comprising:

[0040] a speech recognition module, configured to obtain user speech information and perform speech recognition on the user speech information to obtain a second question description;

[0041] a matching retrieval module, configured to perform a matching retrieval on the second question description using the expert system knowledge base obtained by the expert system knowledge base construction method according to the first aspect, to obtain a second answer content;

[0042] or,

[0043] The model reasoning module is used to reason the second question description through the knowledge reasoning model obtained by the expert system knowledge base construction method as described in the first aspect to obtain the second answer content.

[0044] In a fifth aspect, an embodiment of the present invention provides an expert system knowledge base construction device, comprising:

[0045] at least one processor;

[0046] at least one memory for storing at least one program;

[0047] When the at least one program is executed by the at least one processor, the at least one processor implements the expert system knowledge base construction method described in the first aspect.

[0048] In the sixth aspect, an embodiment of the present invention further provides a computer-readable storage medium, which stores a program executable by a processor, and when the program is executed by the processor, it is used to execute the expert system knowledge base construction method described in the first aspect or the expert system knowledge base question and answer method described in the second aspect.

[0049] The advantages and benefits of the present invention will be described in part in the following description and will become apparent from the following description or learned through practice of the present invention:

[0050] The embodiment of the present invention first obtains a preset first question and answer information, which includes multiple first question descriptions and corresponding question types and first answer content, and then performs feature extraction on the first question description to obtain a question feature vector. Then, based on the question feature vector, the question type and the first answer content, a training data set for deep neural network training is determined, and a knowledge reasoning model is obtained through training, so that expert knowledge data can be obtained based on the knowledge reasoning model, and then an expert system knowledge base can be constructed based on the expert knowledge data. The embodiment of the present invention automatically obtains expert knowledge data through machine learning, thereby improving the construction efficiency and update efficiency of the expert system knowledge base; trains the knowledge reasoning model through a deep neural network, reduces the computing power requirements for the system, and thus reduces the construction cost of the expert system; obtains the user's question description through voice recognition, improves the question and answer communication efficiency of the expert system knowledge base, can be applied and promoted on smart terminal devices, and improves the user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following introduction is made to the drawings required for use in the embodiments of the present invention. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative work.

[0052] Figure 1 A flowchart of a method for constructing an expert system knowledge base provided by an embodiment of the present invention;

[0053] Figure 2 A flowchart of the steps of an expert system knowledge base question-answering method provided by an embodiment of the present invention;

[0054] Figure 3 A structural block diagram of an expert system knowledge base construction system provided by an embodiment of the present invention;

[0055] Figure 4 A structural block diagram of an expert system knowledge base question-answering system provided by an embodiment of the present invention;

[0056] Figure 5 This is a structural block diagram of an expert system knowledge base construction device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and are not to be construed as limiting the present invention. The step numbers in the following embodiments are provided for ease of explanation only and do not limit the order of the steps. The order of execution of the steps in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0058] In the description of the present invention, "a plurality" means two or more. The terms "first" and "second" are used solely to distinguish technical features and are not to be construed as indicating or implying relative importance, or as implicitly indicating the number of the indicated technical features, or as implicitly indicating the order of the indicated technical features. Furthermore, unless otherwise defined, all technical and scientific terms used herein have the same meanings as commonly understood by those skilled in the art.

[0059] Reference Figure 1 The embodiment of the present invention provides a method for constructing an expert system knowledge base, which specifically includes the following steps:

[0060] S101. Obtain preset first question and answer information, where the first question and answer information includes a plurality of first question descriptions and question types and first answer contents corresponding to the first question descriptions.

[0061] Specifically, the embodiment of the present invention organizes the questions and corresponding answers in a prescribed format to form a first question-and-answer message. The first question-and-answer message includes not only the question type and answer content, but also the question stem, the question similarity threshold, supplementary questioning, and intent priority.

[0062] S102: Extract features from the first problem description to obtain a problem feature vector.

[0063] Specifically, the embodiment of the present invention performs word segmentation processing on the first question description, and then obtains a question feature vector based on the word embedding vector of each word. The question feature vector can be used for subsequent model training. Step S102 specifically includes the following steps:

[0064] S1021. Perform word segmentation processing on the first question description to obtain a first phrase;

[0065] S1022. Determine a word embedding vector for each word in the first phrase using a Word2ve model, and then determine a phrase vector for the first phrase based on the word embedding vectors.

[0066] S1023. Normalize the phrase vector to obtain a question feature vector.

[0067] Specifically, an embodiment of the present invention performs word segmentation processing on each problem description to obtain the phrases that make up the problem description. For example, if the problem description is "How to quickly query past bills", after word segmentation processing, the phrase "How to, quickly, query, past, bills" can be obtained. Here, there are many word segmentation algorithms that can be used. For example, in some embodiments, a dictionary-based word segmentation algorithm can be used to first divide the problem description into words according to the dictionary, and then find the best combination of words; in some embodiments, a character-based word segmentation algorithm can also be used to first divide the sentence into individual characters, and then combine the characters into words to find the best combination. After the problem description is word segmented, the word embedding vector corresponding to each word in the phrase can be determined by a pre-established Word2ve model. Of course, in some embodiments, the word embedding vector can be obtained by mapping the word to a vector space with a unified lower dimension. Strategies for generating such mappings include neural networks, dimensionality reduction of word co-occurrence matrices, probabilistic models, and interpretable knowledge base methods. Taking the problem statement "How to quickly query past bills" as an example, first determine the word embedding vector corresponding to each word in the problem statement. The word vector corresponding to the word "how" is (0, 5, 1, 1), the word vector corresponding to the word "quickly" is (0, 0, 0, 1), the word vector corresponding to the word "query" is (4, 2, 3, 1), the word vector corresponding to the word "past" is (0, 1, 0, 1), and the word vector corresponding to the word "bill" is (1, 0, 0, 4). After determining the word embedding vector corresponding to each word in the phrase "how, quickly, query, past, bills," these word embedding vectors can be accumulated and recorded as the phrase vector. The phrase vector corresponding to the phrase "how, quickly, query, past, bills" is (5, 8, 4, 8). Normalizing the phrase vectors yields the problem feature vector corresponding to the problem statement. During normalization, the sum of the elements in the vector corresponding to the phrase can be set to 1. Then, the question description "How to quickly query past bills" can be represented by the vector (0.2, 0.32, 0.16, 0.32). It can be understood that, referring to the above method, the question feature vectors corresponding to all first question descriptions can be determined.

[0068] S103. Determine a training data set based on the question feature vector, question type, and first answer content, and then input the training data set into a pre-built deep neural network for training to obtain a trained knowledge reasoning model.

[0069] As a further optional implementation, the step of determining the training data set based on the question feature vector, the question type, and the first answer content specifically includes:

[0070] A1. Determine the training samples based on the problem feature vector;

[0071] A2. Determine the first-level label based on the question type, and the second-level label based on the content of the first answer;

[0072] A3. Determine the training data set based on the primary labels and secondary labels corresponding to the training samples.

[0073] Specifically, in an embodiment of the present invention, the question type corresponding to the first question description is used as a first-level label, and the first answer content corresponding to the first question description is used as a second-level label, thereby forming a training data set for subsequent multi-label training. The knowledge reasoning model obtained by such training can not only identify the question type of the question description, but also output the corresponding answer content, which facilitates the subsequent classification of expert knowledge data and the construction of the expert system knowledge base.

[0074] As an optional implementation, the step of inputting the training dataset into a pre-built deep neural network for training specifically includes:

[0075] B1. Input the training dataset into the deep neural network to obtain the question type prediction results and the answer content prediction results;

[0076] B2. Determine the training loss value based on the question type prediction result, answer content prediction result, primary label, and secondary label;

[0077] B3. Update the parameters of the deep neural network based on the loss value.

[0078] Specifically, after inputting the data in the training data set into the initialized deep neural network, the prediction results of the model output can be obtained, namely the question type prediction results and the answer content prediction results. The accuracy of the model prediction can be evaluated based on the question type prediction results, the answer content prediction results and the aforementioned primary labels and secondary labels, thereby updating the parameters of the model. In an embodiment of the present invention, the accuracy of the model prediction results can be measured by a loss function (LossFunction). The loss function is defined on a single training data and is used to measure the prediction error of a training data. Specifically, the loss value of the training data is determined by the label of a single training data and the prediction result of the model for the training data. During actual training, a training data set has a lot of training data, so a cost function (CostFunction) is generally used to measure the overall error of the training data set. The cost function is defined on the entire training data set and is used to calculate the average value of the prediction error of all training data, which can better measure the prediction effect of the model. For general machine learning models, the aforementioned cost function, plus a regularization term that measures the complexity of the model, can be used as the objective function of training. Based on this objective function, the loss value of the entire training data set can be calculated. There are many types of commonly used loss functions, such as 0-1 loss function, square loss function, absolute loss function, logarithmic loss function, cross entropy loss function, etc., which can all be used as loss functions of machine learning models, which will not be elaborated here one by one. In an embodiment of the present invention, any one of the loss functions can be selected to determine the loss value of training. Based on the loss value of training, the back propagation algorithm is used to update the parameters of the model, and a trained knowledge reasoning model can be obtained by iterating several rounds. The specific number of iterations can be pre-set, or the training is considered to be completed when the test set meets the accuracy requirements.

[0079] S104. Periodically acquire expert knowledge data according to the knowledge reasoning model, and then construct an expert system knowledge base based on the expert knowledge data.

[0080] Specifically, in the prior art, knowledge engineers often obtain knowledge data from human experts, encode it, and then add it to the expert system knowledge base. However, this method requires collecting a large number of samples, which is inefficient. The embodiments of the present invention can automatically obtain expert knowledge data through a trained knowledge reasoning model, thereby improving the efficiency of building the expert system knowledge base. Step S104 specifically includes the following steps:

[0081] S1041. Obtain a third problem description at a preset period, and reason and classify the third problem description according to the knowledge reasoning model to obtain expert knowledge data and knowledge types of the expert knowledge data;

[0082] S1042. Construct or update the expert system knowledge base based on expert knowledge data and knowledge types.

[0083] Specifically, a cycle is set through a trigger, and the third question description is obtained from the question database of the corresponding field within the cycle. The third question description is then classified into question types and the answer content is inferred through a knowledge reasoning model. The obtained answer content and the question description are then used as expert knowledge data, and the knowledge type is determined according to the question type. The expert system knowledge base can then be constructed or updated based on the expert knowledge data and knowledge type.

[0084] It is understandable that when the above operations are performed for the first time, the expert system knowledge base can be completed. Subsequently, the above steps can be repeated according to a preset cycle to update the expert system knowledge base to expand its content.

[0085] In one embodiment of the present invention, the construction and update of the expert system knowledge base can be completed through the open source WeChat dialogue open platform. The specific process is: the expert knowledge data obtained above is imported through the WeChat dialogue open platform; the entered expert knowledge data is verified on the WeChat dialogue open platform, and the construction of the expert system knowledge base can be started after confirmation; after the construction is completed, the corresponding API interface can be generated to facilitate the subsequent call of the expert system knowledge base.

[0086] The above describes the method for constructing an expert system knowledge base according to an embodiment of the present invention. It can be understood that the embodiment of the present invention automatically acquires expert knowledge data through machine learning, thereby improving the efficiency of constructing and updating the expert system knowledge base; and training the knowledge reasoning model through a deep neural network reduces the computing power requirements of the system, thereby reducing the cost of constructing the expert system.

[0087] Reference Figure 2 The embodiment of the present invention provides an expert system knowledge base question answering method, which specifically includes the following steps:

[0088] S201: Acquire user voice information and perform voice recognition on the user voice information to obtain a second question description;

[0089] S202: performing a matching search on the second question description using the expert system knowledge base obtained by the aforementioned expert system knowledge base construction method to obtain a second answer content;

[0090] or,

[0091] S203. Reason the second question description using the knowledge reasoning model obtained by the aforementioned expert system knowledge base construction method to obtain a second answer content.

[0092] Specifically, taking a mobile terminal device as an example of a human-computer interaction device, user voice information is collected through the voice collection device (such as a microphone, a pickup) on the terminal device; the user's question is identified through the voice recognition algorithm built into the terminal device, converted into text, and a second question description is obtained; the question description is used as a parameter to call the API interface generated by the WeChat dialogue open platform. To ensure the security of user data, an encryption algorithm can be used to encrypt user identity information and question content when calling the interface. For example, user questions and identity information are encapsulated in the XML data packet format and encrypted and decrypted using the PKCS7 algorithm; the reply content returned by the interface is displayed on the terminal device.

[0093] It can be understood that when the second question description can match the corresponding expert knowledge data in the constructed expert system knowledge base, the second answer content can be directly output; when there is no match or the matching degree is too low, the second question description can be model inferenced through the knowledge reasoning model to obtain the second answer content, and at the same time, this group of data can be added to the expert system knowledge base.

[0094] The embodiment of the present invention obtains the user's question description through voice recognition, improves the question-answering communication efficiency of the expert system knowledge base, can be applied and promoted on smart terminal devices, and improves the user experience.

[0095] It should be recognized that the expert system knowledge base construction method and question-answering method provided by the embodiments of the present invention provide a new way for small and medium-sized enterprises and individuals to build industry knowledge bases, greatly reducing the threshold and cost, and improving the user interaction experience. The embodiments of the present invention can be applied to scenarios such as education and teaching, intelligent customer service, and artificial intelligence data annotation. Compared with existing technologies, the embodiments of the present invention also have the following advantages:

[0096] 1) Reduced costs: The embodiments of the present invention can be implemented through mature third-party open source platforms, greatly reducing technical and hardware costs;

[0097] 2) Better experience: Voice interaction makes question-answering more efficient, facilitating the promotion and application of the expert system knowledge base on mobile devices.

[0098] It can be understood that the contents of the above-mentioned expert system knowledge base construction method embodiment are all applicable to the embodiment of the expert system knowledge base question and answer method. The functions specifically implemented by the embodiment of the expert system knowledge base question and answer method are the same as those of the above-mentioned expert system knowledge base construction method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned expert system knowledge base construction method embodiment.

[0099] Reference Figure 3 , an embodiment of the present invention provides an expert system knowledge base construction system, comprising:

[0100] A question and answer information acquisition module is used to acquire preset first question and answer information, where the first question and answer information includes a plurality of first question descriptions and question types and first answer contents corresponding to the first question descriptions;

[0101] A feature extraction module, configured to extract features from the first problem description to obtain a problem feature vector;

[0102] The model training module is used to determine the training data set based on the question feature vector, question type, and first answer content, and then input the training data set into a pre-built deep neural network for training to obtain a trained knowledge reasoning model;

[0103] The knowledge base construction module is used to periodically obtain expert knowledge data according to the knowledge reasoning model, and then construct the expert system knowledge base based on the expert knowledge data.

[0104] It can be understood that the contents of the above-mentioned expert system knowledge base construction method embodiment are all applicable to the present expert system knowledge base construction system embodiment, the functions specifically implemented by the present expert system knowledge base construction system embodiment are the same as those of the above-mentioned expert system knowledge base construction method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned expert system knowledge base construction method embodiment.

[0105] Reference Figure 4 , an embodiment of the present invention provides an expert system knowledge base question answering system, comprising:

[0106] A speech recognition module is used to obtain user speech information and perform speech recognition on the user speech information to obtain a second problem description;

[0107] A matching and retrieval module, configured to perform a matching search on the second question description using the expert system knowledge base obtained by the aforementioned expert system knowledge base construction method to obtain a second answer content;

[0108] or,

[0109] The model reasoning module is used to reason the second question description through the knowledge reasoning model obtained by the aforementioned expert system knowledge base construction method to obtain the second answer content.

[0110] It can be understood that the contents of the above-mentioned expert system knowledge base construction method embodiment are all applicable to the embodiment of the expert system knowledge base question and answer system. The functions specifically implemented by the embodiment of the expert system knowledge base question and answer system are the same as those in the above-mentioned expert system knowledge base construction method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned expert system knowledge base construction method embodiment.

[0111] Reference Figure 5, an embodiment of the present invention provides an expert system knowledge base construction device, comprising:

[0112] at least one processor;

[0113] at least one memory for storing at least one program;

[0114] When the at least one program is executed by the at least one processor, the at least one processor implements the aforementioned method for constructing an expert system knowledge base.

[0115] It can be understood that the contents of the above-mentioned expert system knowledge base construction method embodiment are all applicable to the embodiment of the expert system knowledge base construction device. The functions specifically implemented by the embodiment of the expert system knowledge base construction device are the same as those of the above-mentioned expert system knowledge base construction method embodiment, and the beneficial effects achieved are also the same as those achieved by the above-mentioned expert system knowledge base construction method embodiment.

[0116] An embodiment of the present invention further provides a computer-readable storage medium storing a processor-executable program, which, when executed by the processor, is used to execute the above-mentioned expert system knowledge base construction method or the above-mentioned expert system knowledge base question-answering method.

[0117] A computer-readable storage medium according to an embodiment of the present invention can execute the expert system knowledge base construction method or the expert system knowledge base question-answering method provided by the embodiment of the method of the present invention, can execute any combination of implementation steps of the embodiment of the method of the present invention, and has the corresponding functions and beneficial effects of the embodiment of the method of the present invention.

[0118] The embodiment of the present invention also discloses a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs Figure 1 or Figure 2 The method shown.

[0119] In some optional embodiments, the function / operation mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the function / operation involved, the two boxes shown in succession can actually be executed substantially simultaneously or the above-mentioned boxes can sometimes be executed in reverse order. In addition, the embodiment presented and described in the flow chart of the present invention is provided in an exemplary manner for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operation and logic flow presented herein. Optional embodiments are contemplated in which the order of the various operations is changed and the sub-operations described as a part of a larger operation are performed independently.

[0120] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the above-mentioned functions and / or features can be integrated into a single physical device and / or software module, or one or more functions and / or features can be implemented in separate physical devices or software modules. It is also understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present invention. More specifically, given the properties, functions, and internal relationships of the various functional modules in the devices disclosed herein, the actual implementation of the module will be understood within the routine skills of an engineer. Therefore, a person skilled in the art can implement the present invention set forth in the claims using ordinary skills without undue experimentation. It is also understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, which is determined by the full scope of the appended claims and their equivalents.

[0121] If the above functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the above methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0122] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0123] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable media on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0124] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0125] In the above description of this specification, reference to the terms "one embodiment / example," "another embodiment / example," or "certain embodiments / examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0126] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the claims and their equivalents.

[0127] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A method for constructing an expert system knowledge base, characterized in that: The following steps are involved: Obtaining preset first question-and-answer information, where the first question-and-answer information includes a plurality of first question descriptions and question types and first answer contents corresponding to the first question descriptions; Performing feature extraction on the first problem description to obtain a problem feature vector; Determining a training data set based on the question feature vector, the question type, and the first answer content, and then inputting the training data set into a pre-built deep neural network for training to obtain a trained knowledge reasoning model; periodically acquiring expert knowledge data according to the knowledge reasoning model, and then constructing an expert system knowledge base according to the expert knowledge data; The step of determining a training data set according to the question feature vector, the question type, and the first answer content specifically includes: Determine a training sample according to the problem feature vector; Determine a primary label based on the question type, and determine a secondary label based on the first answer content; Determine a training data set according to the primary labels and secondary labels corresponding to the training samples; The step of inputting the training data set into a pre-built deep neural network for training specifically includes: Inputting the training data set into the deep neural network to obtain question type prediction results and answer content prediction results; Determining a training loss value based on the question type prediction result, the answer content prediction result, the primary label, and the secondary label; The parameters of the deep neural network are updated according to the loss value.

2. The method for constructing an expert system knowledge base according to claim 1, wherein: The step of extracting features from the first problem description to obtain a problem feature vector specifically includes: Performing word segmentation on the first problem description to obtain a first phrase; Determine a word embedding vector for each word in the first phrase using a Word2ve model, and then determine a phrase vector for the first phrase based on the word embedding vector; The phrase vector is normalized to obtain the question feature vector.

3. The method for constructing an expert system knowledge base according to claim 1 or 2, characterized in that: The step of periodically acquiring expert knowledge data according to the knowledge reasoning model and then constructing an expert system knowledge base according to the expert knowledge data specifically includes: Acquire a third problem description at a preset period, and reason and classify the third problem description according to the knowledge reasoning model to obtain expert knowledge data and a knowledge type of the expert knowledge data; An expert system knowledge base is constructed or updated according to the expert knowledge data and the knowledge type.

4. An expert system knowledge base question answering method, characterized in that: The following steps are involved: Acquiring user voice information and performing voice recognition on the user voice information to obtain a second question description; performing a matching search on the second question description using the expert system knowledge base obtained by the expert system knowledge base construction method according to any one of claims 1 to 3 to obtain a second answer content; or, The second question description is inferred by the knowledge reasoning model obtained by the expert system knowledge base construction method according to any one of claims 1 to 3 to obtain the second answer content.

5. An expert system knowledge base construction system, characterized in that: include: A question and answer information acquisition module is used to acquire preset first question and answer information, where the first question and answer information includes a plurality of first question descriptions and question types and first answer contents corresponding to the first question descriptions; A feature extraction module, configured to extract features from the first problem description to obtain a problem feature vector; a model training module, configured to determine a training data set based on the question feature vector, the question type, and the first answer content, and then input the training data set into a pre-built deep neural network for training to obtain a trained knowledge reasoning model; A knowledge base construction module, configured to periodically acquire expert knowledge data according to the knowledge reasoning model, and then construct an expert system knowledge base according to the expert knowledge data; The determining of the training data set according to the question feature vector, the question type, and the first answer content specifically includes: Determine a training sample according to the problem feature vector; Determine a primary label based on the question type, and determine a secondary label based on the first answer content; Determine a training data set according to the primary labels and secondary labels corresponding to the training samples; Inputting the training data set into a pre-built deep neural network for training specifically includes: Inputting the training data set into the deep neural network to obtain question type prediction results and answer content prediction results; Determining a training loss value based on the question type prediction result, the answer content prediction result, the primary label, and the secondary label; The parameters of the deep neural network are updated according to the loss value.

6. An expert system knowledge base question answering system, characterized in that: include: a speech recognition module, configured to obtain user speech information and perform speech recognition on the user speech information to obtain a second question description; a matching retrieval module, configured to perform a matching retrieval on the second question description using the expert system knowledge base obtained by the expert system knowledge base construction method according to any one of claims 1 to 3, to obtain a second answer content; or, The model reasoning module is used to reason the second question description using the knowledge reasoning model obtained by the expert system knowledge base construction method according to any one of claims 1 to 3 to obtain a second answer content.

7. An expert system knowledge base construction device, characterized in that: include: at least one processor; at least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the expert system knowledge base construction method according to any one of claims 1 to 3.

8. A computer-readable storage medium storing a program executable by a processor, characterized in that: The processor-executable program is used to perform the method according to any one of claims 1 to 4 when executed by the processor.

Images