CN115858750A - Power grid technical standard intelligent question-answering method and system based on natural language processing - Google Patents
Power grid technical standard intelligent question-answering method and system based on natural language processing Download PDFInfo
- Publication number
- CN115858750A CN115858750A CN202211507458.7A CN202211507458A CN115858750A CN 115858750 A CN115858750 A CN 115858750A CN 202211507458 A CN202211507458 A CN 202211507458A CN 115858750 A CN115858750 A CN 115858750A
- Authority
- CN
- China
- Prior art keywords
- question
- natural language
- power grid
- answer
- grid technology
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000003058 natural language processing Methods 0.000 title claims abstract description 27
- 238000005516 engineering process Methods 0.000 claims abstract description 51
- 238000010276 construction Methods 0.000 claims abstract description 19
- 238000012549 training Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 13
- 238000005065 mining Methods 0.000 claims description 9
- 230000006870 function Effects 0.000 claims description 8
- 238000003860 storage Methods 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 4
- 238000010586 diagram Methods 0.000 description 9
- 238000012545 processing Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 4
- 238000005315 distribution function Methods 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention belongs to the technical field of question-answering system construction, and provides a power grid technical standard intelligent question-answering method and system based on natural language processing, wherein the method comprises the steps of obtaining power grid technology related natural language questions, searching and constructing a candidate question set similar to the power grid technology related natural language questions from a preset common question answer library; calculating the similarity between the natural language question related to the power grid technology and each statement in the candidate question set, and if the similarity is greater than a first similarity threshold value, outputting an answer corresponding to the corresponding statement in the candidate question set; and if the similarity is smaller than the second similarity threshold, obtaining an answer corresponding to the natural language question related to the power grid technology based on the reading comprehension model.
Description
Technical Field
The invention belongs to the technical field of question-answering system construction, and particularly relates to a power grid technical standard intelligent question-answering method and system based on natural language processing.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
Under the global era background of Internet interviews, intelligent natural language question answering is an important means for processing language information, and intelligent question answering which is efficient, comprehensive and more practical is one of key research directions of natural language intelligence. Due to professional limitation of electric power industry terminology, the current universal question-answering community module cannot be well adapted to the electric power industry terminology, accuracy of an electric power question-answering model is seriously influenced, and user experience is poor. In addition, the question-answering system cannot meet the increasing demand of the user for question-answering.
Disclosure of Invention
In order to solve at least one technical problem existing in the background technology, the invention provides a power grid technical standard intelligent question-answering method and system based on natural language processing, which aim to build technical standard question-answering system models based on different fields of the power industry, can recognize more intelligent sentences from the aspects of semantics, contexts and language structures, and aim to build a more practical question-answering system.
In order to achieve the purpose, the invention adopts the following technical scheme:
the invention provides a power grid technical standard intelligent question-answering method based on natural language processing.
A power grid technical standard intelligent question-answering method based on natural language processing comprises the following steps:
acquiring a natural language problem related to a power grid technology, and searching and constructing a candidate problem set similar to the natural language problem related to the power grid technology from a preset common problem answer library; the preset common question answer library is stored with a plurality of power grid technology related questions and corresponding answers thereof in advance;
calculating the similarity between the power grid technology-related natural language question and each statement in the candidate question set, and if the similarity is greater than a first similarity threshold value, outputting an answer corresponding to the corresponding statement in the candidate question set; if the similarity is smaller than a second similarity threshold value, obtaining answers corresponding to the power grid technology related natural language questions based on a reading comprehension model;
wherein the first similarity threshold is greater than the second similarity threshold; the reading and understanding type model is constructed by the following steps: based on the construction mode of question-answer-clause, a prompt learning method and a power clause question-answer data set are adopted for training, and answers of the questions are generated by mining semantic relations between the questions and the clauses.
As an embodiment, the question-answer-clause construction is obtained by constructing a complete filling template and a continuation character string prefix.
As an embodiment, the prompt learning method introduces additional parameters and uses an objective function of a preset task to perform fine tuning on the reading comprehension model so as to pre-train the reading comprehension model to adapt to different downstream tasks.
As an embodiment, for each answer type, the prompt learning method needs to define a new answer set, where all answer labels in the set are words in the pre-training model;
each answer type corresponds to a plurality of new label sets, and a whole dictionary set is finally obtained by taking a union set of each label.
In one embodiment, the categories of the terms include a number category, an extraction category, a statistics category, and a judgment category.
In one embodiment, in the mining of the semantic relationship between the questions and the terms, based on an existing power system question-answer dataset, samples in the dataset are represented, and a sample format of each dataset is as follows:
x＝{[cls],t 1 ,t 2 ,...,m,...,t T ,[sep]}
wherein, [ cls ]]Is a sentence prefix identifier; [ sep ]]A separator mark between different sentences; t is t 1 ,t 2 ,...,m,...,t T A representation for each word; m is the portion of the mask language model mask, i.e., the location where the model needs to learn and predict.
As an embodiment, the similarity between the grid technology related natural language question and each statement in the candidate question set is measured by cosine similarity or euclidean distance.
The invention provides a power grid technical standard intelligent question-answering system based on natural language processing.
A power grid technical standard intelligent question-answering system based on natural language processing comprises:
the candidate problem set construction module is used for acquiring the natural language problems related to the power grid technology, searching and constructing a candidate problem set similar to the natural language problems related to the power grid technology from a preset common problem answer library; the preset common question answer library is stored with a plurality of power grid technology related questions and corresponding answers thereof in advance;
the question answer obtaining module is used for calculating the similarity between the power grid technology-related natural language question and each statement in the candidate question set, and if the similarity is greater than a first similarity threshold value, outputting an answer corresponding to the corresponding statement in the candidate question set; if the similarity is smaller than a second similarity threshold value, obtaining answers corresponding to the natural language questions related to the power grid technology based on a reading comprehension model;
wherein the first similarity threshold is greater than the second similarity threshold; the construction process of the reading and understanding type model comprises the following steps: based on the construction mode of question-answer-clause, a prompt learning method and a power clause question-answer data set are adopted for training, and answers of the questions are generated by mining semantic relations between the questions and the clauses.
A third aspect of the invention provides a computer-readable storage medium.
A computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the grid technology standard intelligent question-answering method based on natural language processing as described above.
A fourth aspect of the invention provides a computer apparatus.
A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the grid technology standard intelligent question answering method based on natural language processing.
Compared with the prior art, the invention has the beneficial effects that:
aiming at the self term characteristics and semantic specifications in the power industry, the invention combines the technical achievement of the front edge of the natural language processing field to realize the accurate positioning of the problems and the accurate generation of answers, integrates different deep learning models such as an intention recognition model, the training and construction of a semantic vector pair matching retrieval model, a reading and understanding type question-answering model and the like, integrates different algorithms such as integrated learning, less-sample learning, larger-scale context learning, query reconstruction, question-answering-based multi-task learning, controllable generation, attention supervision, data enhancement and the like, and accurately positions and accurately analyzes the task target in different stages of the construction of the question-answering system.
Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
Fig. 1 is a flow diagram of an intelligent question-answering method based on natural language processing for power grid technology standards according to the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
The invention mainly aims to build a question-answering system in the power industry, and the question-answering model system in the power industry is built by using the related technology in the field of deep learning-natural language processing aiming at different question-answering situations of extraction, judgment, digital and statistic classes on the basis of the accumulation of a clause question-answering corpus of the power industry and the learning of the semantic and context information of the power corpus, so that a quick solution is provided for the discovery and processing of the related problems in the power field.
Example one
As shown in fig. 1, the present embodiment provides a power grid technology standard intelligent question-answering method based on natural language processing, which includes the following steps:
step 1: acquiring natural language problems related to the power grid technology, and searching and constructing a candidate problem set similar to the natural language problems related to the power grid technology from a preset frequently-used problem answer library; the preset common problem answer library is stored with a plurality of power grid technology related problems and corresponding answers thereof in advance.
Step 2: calculating the similarity between the natural language question related to the power grid technology and each statement in the candidate question set, and if the similarity is greater than a first similarity threshold value, outputting an answer corresponding to the corresponding statement in the candidate question set; if the similarity is smaller than a second similarity threshold value, obtaining answers corresponding to the power grid technology related natural language questions based on a reading comprehension model;
wherein the first similarity threshold is greater than the second similarity threshold; the reading and understanding type model is constructed by the following steps: based on the construction mode of question-answer-clause, a prompt learning method and a power clause question-answer data set are adopted for training, and answers of the questions are generated by mining semantic relations between the questions and the clauses.
For example, the first similarity threshold may be 0.85, and the second similarity threshold may be 0.4.
In one or more embodiments, the question-answer-clause construction is achieved by constructing a complete fill-in template and extending a prefix of a string.
And constructing and training a question-answer-clause matching model for the reading comprehension type question-answering system in a mode of constructing a complete blank-filling template and extending the prefix of the character string.
The reading and understanding model adopts a bert model in the deep learning field, and sample data is fitted in a prompt training mode.
Specifically, the reading-understanding model utilizes a prompt learning method based on a Mask Language Model (MLM), retrains the model again based on downstream tasks such as text classification, reading understanding and the like by introducing a data template, applies the pre-trained language model to a natural language processing task of the power industry, and enables the bert model trained by the open corpus to accurately understand text content and language environment of the power industry.
As one or more embodiments, extracting key semantic information after the question input by the user specifically includes:
and processing and extracting key semantic information in the problems and the clauses through a unified vector generation model to respectively obtain a question vector and a clause vector to form a unified feature vector of the problems and the clauses.
In step 2, as one or more embodiments, in mining the semantic relationship between the question and the term:
(1) Database for answering based on existing power system 1 ,...,x n Represents the samples in the dataset as:
the samples in the dataset are composed in the manner of x i And = question-answer-sample concatenation. x is the number of i = i sample.
(2) A sentence is input, i.e. a sample format of the data set is: x =: [ cls],t 1 ,2,...,,..., T ,[sep]}
Wherein, [ cls ]]Is a period head identifier, [ sep]Is a delimiter mark between different sentences, [ cls ]]And [ sep ]]Middle t i For each word representation, m is the portion of the mask language model mask, i.e., the location where the model needs to learn and predict.
(3) Obtaining latent layer (hidden layer) containing semanteme through pre-training, { h [cls] ,h 1 ,...,h T ,h [sep] }；
(4) Finally choose as h [cls] Last hidden layerAnd expressing that the prediction effect is achieved by mapping the neural network weight matrix W and the softmax activation function to probability distribution, wherein the final result of softmax is the probability that the answer belongs to different answers, and the category with the maximum probability is the final answer to the question.
Wherein the probability distribution function is:
P(y∈Y|s)＝softmax(Wh [cls] +b)
wherein W is a weight matrix, b is an offset, and represents the weight matrix of W and h [cls] The hidden layer represents the bias of the matrix inner product result. s is a question, Y is a set of answers, and Y is one answer in the set of answers.
And in the training process, wherein W, b and all pre-training model parameters are adjusted through loss, wherein the loss function is as follows:
where n is the number of questions, s i Is the ith question, y i Is the ith answer; p (-) represents a probability distribution function.
For each answer type, the prompt learning method needs to define a new answer set, where all answers label in the set are words in the pre-training model vocab:
v y ＝{w 1 ,...,w m }
each answer type corresponds to a plurality of new label sets, and a whole dictionary set T(s) is finally obtained by taking the union set of each label.
P(y∈Y|s)＝P([mask]＝w∈v y |T(s))
According to the scheme, a Mask Language Model (MLM) is applied to the question and answer fields of the extraction type, the judgment type, the number type and the statistic type clause type for training through a question-answer-clause construction mode, answers are learned through semantic relations between questions and clauses, and the generation model has the capability of generating answers according to the questions and the clauses.
In this embodiment, the key semantic information in the question and the clause is processed and extracted through a unified vector generation model, so as to form a unified feature vector expression manner of the question and the clause (that is, the original question vector generation model and the original clause vector generation model in the figure are combined into one, and similar feature expressions are formed for similar semantic relationships in the question and the clause). In the face of the generation of a new question, the model first receives a question posed by the user (i.e., a question input by the user). Then, according to the question sentence input by the user, a candidate question set which is similar to the user question is searched and established from the common question library. And then, cosine similarity calculation is carried out on the sentences in the candidate question set and the question input by the user, and the question most similar to the question input by the user is searched from the candidate question set. If a question similar to the question input by the user is found in the candidate question set (namely, the similarity between the question and the question input by the user is greater than a certain threshold), directly returning the answer corresponding to the question to the user; if a question similar to the question input by the user is not found (namely, the similarity between all the questions in the candidate question set and the question input by the user is smaller than a certain threshold value), the retrieval of the relevant questions and the giving of the answers are carried out by using the keyword matching principle, and the new question and the corresponding answer are added into the question-answer library.
Example two
This embodiment provides electric wire netting technical standard intelligence question-answering system based on natural language is handled, includes:
the candidate question set construction module is used for acquiring natural language questions related to the power grid technology, searching and constructing a candidate question set similar to the natural language questions related to the power grid technology from a preset common question answer library; the preset common question answer library is stored with a plurality of power grid technology related questions and corresponding answers thereof in advance;
the question answer obtaining module is used for calculating the similarity between the power grid technology-related natural language question and each statement in the candidate question set, and if the similarity is greater than a first similarity threshold value, outputting an answer corresponding to the corresponding statement in the candidate question set; if the similarity is smaller than a second similarity threshold value, obtaining answers corresponding to the natural language questions related to the power grid technology based on a reading comprehension model;
wherein the first similarity threshold is greater than the second similarity threshold; the construction process of the reading and understanding type model comprises the following steps: based on the construction mode of question-answer-clause, a prompt learning method and a power clause question-answer data set are adopted for training, and answers of the questions are generated by mining semantic relations between the questions and the clauses.
It should be noted that, each module in the present embodiment corresponds to each step in the first embodiment one to one, and the specific implementation process is the same, and is not described in detail here.
EXAMPLE III
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the natural language processing based grid technology standard intelligent question answering method as described above.
Example four
The embodiment provides a computer device, which includes a memory, a processor and a computer program stored in the memory and executable on the processor, and when the processor executes the program, the steps in the natural language processing-based power grid technology standard intelligent question answering method described above are implemented.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. The power grid technical standard intelligent question-answering method based on natural language processing is characterized by comprising the following steps of:
acquiring natural language problems related to the power grid technology, and searching and constructing a candidate problem set similar to the natural language problems related to the power grid technology from a preset frequently-used problem answer library; the power grid technology-related questions and the corresponding answers thereof are stored in the preset frequently-used question answer library in advance;
calculating the similarity between the natural language question related to the power grid technology and each statement in the candidate question set, and if the similarity is greater than a first similarity threshold value, outputting an answer corresponding to the corresponding statement in the candidate question set; if the similarity is smaller than a second similarity threshold value, obtaining answers corresponding to the power grid technology related natural language questions based on a reading comprehension model;
wherein the first similarity threshold is greater than the second similarity threshold; the reading and understanding type model is constructed by the following steps: based on the construction mode of question-answer-clause, a prompt learning method and a power clause question-answer data set are adopted for training, and answers of the questions are generated by mining semantic relations between the questions and the clauses.
2. The grid technology standard intelligent question-answering method based on natural language processing as claimed in claim 1, wherein the construction of the question-answer-clause is obtained by constructing a complete gap-filling template and a character string prefix.
3. The natural language processing-based power grid technical standard intelligent question-answering method as claimed in claim 1, wherein the prompt learning method is used for pre-training the reading comprehension model to adapt to different downstream tasks by introducing additional parameters and using an objective function of a preset task to perform fine adjustment on the reading comprehension model.
4. The natural language processing-based power grid technology standard intelligent question-answering method according to claim 3, wherein for each answer type, the prompt learning method needs to define a new answer set, and all answer labels in the set are words in a pre-training model;
each answer type corresponds to a plurality of new label sets, and a whole dictionary set is finally obtained by taking a union set of each label.
5. A natural language processing based power grid technology standard intelligent question answering method according to claim 1, wherein the categories of the terms include a number category, an extraction category, a statistics category and a judgment category.
6. The natural language processing-based power grid technology standard intelligent question-answering method according to claim 1, wherein in the mining of semantic relationships between questions and terms, samples in data sets are represented based on existing power system question-answering data sets, and the sample format of each data set is as follows:
x＝{[cls],t 1 ,2,...,,..., T ,[sep]}
wherein, [ cls ]]Is a sentence start identifier; [ sep]A separator mark between different sentences; t is t 1 , 2 ,...,,..., T A representation for each word; m is the portion of the mask language model mask, i.e., the location where the model needs to learn and predict.
7. The natural language processing-based power grid technology standard intelligent question-answering method according to claim 1, wherein the similarity between the power grid technology-related natural language question and each statement in the candidate question set is measured by cosine similarity or Euclidean distance.
8. Electric wire netting technical standard intelligence question-answering system based on natural language is handled, its characterized in that includes:
the candidate question set construction module is used for acquiring natural language questions related to the power grid technology, searching and constructing a candidate question set similar to the natural language questions related to the power grid technology from a preset common question answer library; the preset common question answer library is stored with a plurality of power grid technology related questions and corresponding answers thereof in advance;
the question answer obtaining module is used for calculating the similarity between the power grid technology-related natural language question and each statement in the candidate question set, and if the similarity is greater than a first similarity threshold value, outputting an answer corresponding to the corresponding statement in the candidate question set; if the similarity is smaller than a second similarity threshold value, obtaining answers corresponding to the natural language questions related to the power grid technology based on a reading comprehension model;
wherein the first similarity threshold is greater than the second similarity threshold; the reading and understanding type model is constructed by the following steps: based on the construction mode of question-answer-clause, a prompt learning method and a power clause question-answer data set are adopted for training, and answers of the questions are generated by mining semantic relations between the questions and the clauses.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the steps in the natural language processing based grid technology standard intelligent question answering method according to any one of claims 1 to 7.
10. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps in the natural language processing based grid technology standard intelligent question answering method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211507458.7A CN115858750A (en) | 2022-11-29 | 2022-11-29 | Power grid technical standard intelligent question-answering method and system based on natural language processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211507458.7A CN115858750A (en) | 2022-11-29 | 2022-11-29 | Power grid technical standard intelligent question-answering method and system based on natural language processing |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115858750A true CN115858750A (en) | 2023-03-28 |
Family
ID=85667510
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211507458.7A Pending CN115858750A (en) | 2022-11-29 | 2022-11-29 | Power grid technical standard intelligent question-answering method and system based on natural language processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115858750A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116186232A (en) * | 2023-04-26 | 2023-05-30 | 中国电子技术标准化研究院 | Standard knowledge intelligent question-answering implementation method, device, equipment and medium |
| CN117312534A (en) * | 2023-11-28 | 2023-12-29 | 南京中孚信息技术有限公司 | Intelligent question-answering implementation method, device and medium based on secret knowledge base |
-
2022
- 2022-11-29 CN CN202211507458.7A patent/CN115858750A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116186232A (en) * | 2023-04-26 | 2023-05-30 | 中国电子技术标准化研究院 | Standard knowledge intelligent question-answering implementation method, device, equipment and medium |
| CN117312534A (en) * | 2023-11-28 | 2023-12-29 | 南京中孚信息技术有限公司 | Intelligent question-answering implementation method, device and medium based on secret knowledge base |
| CN117312534B (en) * | 2023-11-28 | 2024-02-23 | 南京中孚信息技术有限公司 | Intelligent question-answering implementation method, device and medium based on secret knowledge base |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107798140B (en) | Dialog system construction method, semantic controlled response method and device | |
| CN114064918B (en) | Multi-modal event knowledge graph construction method | |
| CN111666758B (en) | Chinese word segmentation method, training device and computer readable storage medium | |
| CN111274790B (en) | Chapter-level event embedding method and device based on syntactic dependency graph | |
| CN115858750A (en) | Power grid technical standard intelligent question-answering method and system based on natural language processing | |
| CN112541337B (en) | Document template automatic generation method and system based on recurrent neural network language model | |
| CN109062904B (en) | Logic predicate extraction method and device | |
| CN116166782A (en) | Intelligent question-answering method based on deep learning | |
| CN112364132A (en) | Similarity calculation model and system based on dependency syntax and method for building system | |
| CN114676255A (en) | Text processing method, device, equipment, storage medium and computer program product | |
| CN115577086A (en) | Bridge detection knowledge graph question-answering method based on hierarchical cross attention mechanism | |
| CN115759119A (en) | Financial text emotion analysis method, system, medium and equipment | |
| CN111666374A (en) | Method for integrating additional knowledge information into deep language model | |
| CN116522165B (en) | Public opinion text matching system and method based on twin structure | |
| CN110377753B (en) | Relation extraction method and device based on relation trigger word and GRU model | |
| CN117313850A (en) | Information extraction and knowledge graph construction system and method | |
| CN114637852A (en) | Method, device and equipment for extracting entity relationship of medical text and storage medium | |
| CN113449038A (en) | Mine intelligent question-answering system and method based on self-encoder | |
| Li | Construction of English Translation Model Based on Improved Fuzzy Semantic Optimal Control of GLR Algorithm | |
| CN114357166B (en) | Text classification method based on deep learning | |
| CN116757195A (en) | Implicit emotion recognition method based on prompt learning | |
| Yang | Intelligent English Translation Evaluation System Based on Internet Automation Technology | |
| Yang et al. | Question answering system based on Knowledge Graph | |
| CN115359486A (en) | Method and system for determining custom information in document image | |
| CN116822508A (en) | Structured knowledge extraction method based on text understanding |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |