15 results about "Knowledge base question answering" patented technology

This application relates to an energy knowledge base question answering method and an energy knowledge base question answering device. The method includes: receiving an original question containing a role identifier, and determining a job vocabulary corresponding to the role identifier; splicing the lexical units in the original question that match the job vocabulary to obtain a job-limited text, mapping the job-limited text to a predetermined slot structure to obtain an intent lexical unit group; determining the latest approval node in the energy knowledge base that matches the role identifier and the intent lexical unit group; filtering in the energy knowledge base with the latest approval node to obtain a filtered business segment, matching the intent lexical unit group with the filtered business segment weighted by an approval weight, and obtaining a candidate business segment, where the approval weight is determined according to the level of the approval node corresponding to the filtered business segment; screening out a target business segment from the candidate business segments based on the intent lexical unit group; and generating a reply text based on the target business segment and the intent lexical unit group. Using this method can improve the accuracy of the reply.

The application discloses a knowledge base question answering method and system under full-process large language model small sample learning, the method comprises the following steps: question analysis, identifying candidate triples and question types of original user questions through a large language model; query template generation, generating a knowledge base query statement template with placeholders for user questions; knowledge base linking, obtaining matched expressions for the placeholders in the query statement template in combination with a large language model and an entity information index that has been constructed; answer generation, generating a knowledge base query language and returning a final result. Through the small sample learning capability of the large model, the retrieval process is redesigned for the knowledge base question answering task, a model does not need to be trained separately, and the question answering process is simplified; through the powerful knowledge reserve of the large model, the question answering effect and generalization capability are improved.

The application discloses a data retrieval method, a server, a terminal and a storage medium, which are used for providing a knowledge base question and answer retrieval function for VDI users through a server without occupying the storage space of the server, and providing a safer, more reliable and more intelligent question and answer retrieval scheme. The method comprises the following steps: acquiring first text information from a first remote terminal in one or more remote terminals, wherein the first text information corresponds to a first embedding vector; if the similarity of the first embedding vector and a second embedding vector in a vector database is greater than or equal to a threshold value, determining a retrieval result of the first text information according to the second text information corresponding to the second embedding vector and the first embedding vector, wherein the second text information is obtained by text segmentation according to source documents of the one or more remote terminals.

The application discloses a knowledge base question answering method and device, computer equipment and a storage medium. The method comprises the following steps: in response to a user query, extracting a structured query element corresponding to a preset query dimension; the preset query dimension at least comprises a domain keyword; based on the structured query element, searching in a preset knowledge base to obtain a candidate knowledge record set; the preset knowledge base comprises a plurality of knowledge records, and each knowledge record at least comprises an identifier field and a domain keyword field, and each identifier is associated with a semantic vector generated based on the content of the corresponding knowledge record; encoding the user query into a query vector, and performing similarity matching between the query vector and the semantic vector associated with the identifier of each knowledge record in the candidate knowledge record set; and filtering a target knowledge record set from the candidate knowledge record set according to the matching result, and generating an answer corresponding to the user query based on the target knowledge record set. Thus, the accuracy of the answer is improved.

The application discloses an intelligent evaluation system and method for a knowledge base question answering application, and relates to the technical fields of artificial intelligence, natural language processing and multi-agent collaborative system. The system comprises a data preprocessing module, a data synthesis module, a data screening module, a data scoring module and an adaptive weight adjustment module. The data preprocessing module performs data preprocessing on enterprise original documents and generates a data knowledge base. The data synthesis module repeatedly asks and answers each paragraph of the data knowledge base by using a large language model to generate a question and answer pair. The data screening module screens out high-quality samples through screening agent collaboration and a MACA framework mechanism. The data scoring module scores the question and answer pair through scoring agent collaboration and a MACA framework mechanism. The adaptive weight adjustment module dynamically updates the weight of each screening and scoring dimension by using an exponentially weighted moving average algorithm.

The application relates to the technical field of intelligent query of low-altitude flight regulations, and discloses a low-altitude regulation knowledge base question-answering system based on adaptive RAG, which comprises the following modules: a specification document preprocessing module, which is used for data cleaning of low-altitude multi-source heterogeneous regulation documents, extraction of multi-dimensional information, and establishment of a multi-dimensional label system; an adaptive RAG knowledge base construction module, which is used for structured storage of a clause unit and a label system structure and construction of a double-mode knowledge base; an adaptive routing and differentiated retrieval module, which is used for complexity discrimination of user queries and adaptive selection of different retrieval paths; a context integration and generation module, which is used for splicing of retrieval results, source information and user queries after sorting of the retrieval results, and generation of an interpretable answer based on reference materials; and a feedback optimization module, which is used for closed-loop optimization through periodical incremental fine-tuning of the model. The application can deeply understand the space-time constraint characteristics of the low-altitude field, dynamically resolve multi-source regulation conflicts, and realize efficient and accurate intelligent question-answering.

The application relates to the technical field of information retrieval processing, and discloses an aviation knowledge base question answering method and system based on vector retrieval and a large model, which comprises the following steps: analyzing a business document into rule units to generate rule vectors; receiving a question to generate a question vector; judging whether a query condition satisfies necessary conditions; if the necessary conditions are not satisfied, outputting supplementary question information; if the necessary conditions are satisfied, screening the rule units, performing retrieval, and obtaining recalled rule units; judging whether there is a rule conflict among the recalled rule units; if there is a rule conflict, judging whether there is a substitution relationship or an exception relationship; if there is a substitution relationship or an exception relationship, determining an effective rule unit; if there is no substitution relationship or exception relationship, outputting verification information; if there is no rule conflict, determining the recalled rule units as the effective rule units; generating a question and answer result according to the effective rule units; and outputting a final question and answer result. The application improves the accuracy, pertinence and stability of an aviation knowledge base question and answer result.

This application relates to the field of retrieval enhancement and generation technology, and in particular to a method and system for collaborative construction of multi-source knowledge bases and text question answering in a professional domain. The construction method includes acquiring professional domain text data and performing block processing on the professional domain text data; constructing an initial knowledge graph, a text knowledge base, and a question answering knowledge base; binning the entity set in the initial knowledge graph by type, extracting connected components to obtain candidate entity clusters, employing a two-layer scheduling strategy for semantic determination using a large language model, and progressively solidifying the determination results and performing targeted supplementary judgments on unresolved nodes; updating the initial knowledge graph to form a knowledge graph library; and obtaining a multi-source heterogeneous knowledge base. The method and system for collaborative construction of multi-source knowledge bases and text question answering in this professional domain can solve the problems of insufficient quality in vertical domain knowledge graph construction, low entity disambiguation accuracy, and high computational cost, and also features high stability and adaptive adjustment.

The application relates to the technical field of data processing, and discloses a question and answer processing method, device, equipment and medium. Through joint training of a sorting task and an answer type prediction task, a model learns logical type constraints between questions and answers while optimizing semantic correlation, effectively suppresses false correlations caused by training data bias or high-frequency answer interference. Further, in the question and answer reasoning process, with the aid of an answer path sorting model and a cross-sequence interaction attention mechanism, comprehensive sorting results of each candidate answer path fusion semantic score and type consistency score can be obtained, multi-granularity accurate matching from overall semantics to local elements is realized, and the sorting accuracy and model robustness in a complex query scene are significantly improved. Therefore, in the knowledge base question and answer task in the fields of finance and insurance, medical treatment and the like, questions with similar semantics but different logical types can be effectively distinguished, information confusion is avoided, and the final answer has accuracy and reliability.

A computer-implemented method for providing artificial intelligence (AI)-driven interactive knowledge base question-answering based on multiple large-language model (LLM) Iterations with separate software tools for LLM evaluation, the method comprising receiving a query; generating, based on the query, prompts comprising contextual information associated with the query, and a reference to a database comprising associated with the query; initiating a first LLM to generate, based on the prompts and the database, a first response to the query; receiving from the first LLM, the first response; evaluating an accuracy of the first response using at least one software tool separate from the first LLM; initiating, based on the evaluating, a second LLM to generate a final response based on the prompts and the first response from the first LLM; and providing the final response to the query.

The application discloses an education field knowledge base search optimization method based on question generation, which comprises the following steps: firstly, obtaining the education field text by analyzing the education knowledge base direction field corpus; obtaining the semantic model by using the education field text to pre-train the language model migration learning; designing the fixed question and answer pair template based on the existing structured text information in the knowledge base to obtain the knowledge base question and answer pair; training the question generation model by using the knowledge base question and answer pair data and the Chinese open source question and answer pair data, deploying the question generation reasoning service; generating the question and answer pair to expand the knowledge base; simultaneously coding the entity node text in the knowledge base structured information and the question text in the question and answer pair by using the semantic model to construct the vector library, and performing the semantic similarity calculation after the user query input; recalling the best result in the online semantic matching; and the application greatly improves the recall rate of the returned result of the user search behavior, improves the learning efficiency and improves the user experience.

This invention discloses a large-model retrieval enhancement method based on multi-granularity dynamic adaptation, comprising the following steps: S1: Selecting and preprocessing the evaluation dataset; S2: Constructing a vector library with three granularities; S3: Allocating granularity to the input question; S4: Retrieving the TOP-K text blocks related to the question; S5: Selecting the most suitable text block through a budgeter; S6: Packaging all selected text blocks and returning them to the large model for answering. This invention can achieve adaptive coordination of retrieval granularity, candidate size, and final evidence set under different question types and evidence distributions, improving the integrity of the evidence chain and reducing noise interference. While maintaining a simple and easy-to-maintain block partitioning scheme, this invention improves the stability and accuracy of multi-hop question answering, and is suitable for application scenarios such as knowledge base question answering, intelligent retrieval, and question answering services. It can also be extended to tasks such as long document question answering and cross-document information aggregation.

The application discloses a Chinese knowledge base question answering method based on a double processing system. The knowledge base question answering model mainly comprises a coordination perception module (System 1) and an explicit reasoning module (System 2), and each model component is interrelated and closely combined. Under the demand of the existing knowledge base question answering task, the application proposes a Chinese knowledge base question answering model based on the double processing system. Based on the question raised by a user, System 1 firstly learns the representation of the question and predicts the corresponding simple path, then System 2 retrieves the complex path in the knowledge base for the question according to the simple path generated by System 1, and finally the trained model is used to predict the answer. The application is based on artificial intelligence technology and combines the double-process cognitive theory to design a new type of Chinese knowledge base question answering model. The method has wide application prospects in the intelligent question answering field of various industries.

This invention provides an optimization method for a temporally dependent data prediction model based on agent trajectory feedback, applicable to tasks such as webpage evidence text collection, open encyclopedia question answering, enterprise knowledge base question answering, or retrieval enhancement generation. The method collects multi-round execution trajectories formed by the agent's interaction with the retrieval system during task execution. Initial supervised samples are constructed based on the candidate document set returned by the search action and subsequent browsing actions. Positive samples are filtered using post-browsing inference text and an inference-evidence consistency judgment model. Relevance strength weights are estimated based on inference length, evidence citation count, number of fact entries, and changes in subsequent actions. The retrieval model is trained using temporally gated coding and a weighted contrastive learning objective function. The optimized retrieval model is then deployed back to the agent system for continuous closed-loop updates, improving the execution efficiency of webpage evidence collection, evidence retrieval, and complex question answering tasks.

The application provides a drug regulatory industry knowledge base question answering system, method, electronic device and storage medium. The knowledge base question answering system comprises: a knowledge base warehousing module, comprising a question and answer library and a document library; the question and answer library comprises questions and answers, and the document library comprises drug regulatory industry documents; a knowledge base question answering module is configured to determine user input information, and determine first answering information corresponding to the user input information according to the question and answer library; when the first answering information corresponding to the user input information is not determined according to the question and answer library, second answering information corresponding to the user input information is determined according to the document library; and when the second answering information corresponding to the user input information is not determined according to the document library, generative answering is performed. Through the multi-stage and multi-level knowledge base warehousing processing, the knowledge retrieval accuracy and precision of the knowledge base question answering system can be improved.