219 results about "Semantic integrity" patented technology

The invention discloses a hierarchical semantic-driven retrieval enhancement generation method and system, a natural hierarchical relationship and a semantic boundary of a document are effectively reserved by constructing a tree hierarchical structure based on a document chapter title, and a recursive semantic boundary splitting strategy is adopted to refine overlong text nodes, so that the semantic integrity is ensured, and the retrieval enhancement generation efficiency is improved. And the model input length limitation is met, and the semantic information is prevented from being lost. Meanwhile, node knowledge point extraction and abstract generation are achieved through a large language model, top-down multi-level title path transmission and bottom-up content aggregation are combined, the structural perception and semantic expression ability of nodes is enhanced, and in the retrieval stage, based on similarity distribution of query and node semantic expression, an adaptive retrieval threshold value is dynamically calculated, and the retrieval efficiency is improved. A fixed top-k retrieval strategy is replaced, intelligent screening of different query and hierarchical nodes is achieved, information coverage and redundancy suppression are balanced, and retrieval efficiency and accuracy are remarkably improved.

The invention provides a multi-modal intelligent question-answering system based on a large model and a construction method and device, and the method comprises the steps: carrying out the region recognition of a PDF document, and obtaining a table region picture, a text region picture and / or an image region picture; the Word document is analyzed, paragraph content of the Word document is obtained, and the paragraph content is composed of the first text, table data and / or Word pictures; constructing a knowledge base; wherein the knowledge base comprises vector representation of document slices; the document slice is determined based on the table area picture, the character area picture and / or the image area picture, or the document slice is determined based on the paragraph content and the corresponding chapter information thereof. Visibly, according to the method, layout recognition can be carried out on the PDF document, catalog chapter recognition can be carried out on the Word document, the knowledge base is constructed based on the layout recognition, the semantic integrity and coherence of multi-modal text question and answer recall can be improved, and the answer accuracy is improved.

The invention provides a semantic enhancement adaptive partitioning method and system for natural resource large model questions and answers, and aims to solve the problems of difficulty in term boundary recognition, damage to semantic integrity and the like. According to the method, three core technologies including theme perception coarse-grained paragraph division, self-adaptive sliding window theme hierarchy division and embedded perception context self-adaptive text segmentation are fused. The system firstly analyzes a natural resource long text structure, identifies titles and theme levels and aligns associated contents; paragraphs are extracted according to a theme perception strategy and are subdivided into sentence sets according to grammar rules; an improved sliding window mechanism is adopted to divide sentences into window sentence block groups. The method is characterized in that a dynamic aggregation threshold mechanism is introduced, the semantic association degree between adjacent sentence blocks is calculated through an embedded perception context semantic segmentation technology, whether the sentence blocks are combined or not is judged by combining a similarity distribution change trend and a dynamic adjustment threshold, self-adaptive delimitation of semantic boundaries is achieved, and text blocks which are clear in structure and coherent in semantics are generated.

The invention relates to the field of document processing and data extraction, and particularly discloses a PDF text extraction method and system based on a large language model.The method includes the steps that content of all pages of a target PDF document is positioned and marked to obtain a first to-be-recognized area and a second to-be-recognized area, and noise interference features of the to-be-recognized areas are removed; formulating a multi-level text logic reconstruction strategy to complete reconstruction of a logic sequence of the target PDF document, preliminarily outputting a first-level PDF document, and performing primary image-text association degree analysis to output first association strength; performing intelligent anomaly recognition and correction on the content of the first-level PDF document on a semantic structure through a large language model to obtain a second-level PDF document, and outputting second association strength; judging whether the secondary PDF document is qualified or not based on the first association strength and the second association strength; according to the method, the logic sequence and the semantic integrity of the document can be recovered, and the text purity and the structural integrity are improved.

The invention relates to the technical field of data processing, and discloses a data processing method, a data processing model obtaining method, a data processing model obtaining device, data processing equipment and a medium. According to the data processing model obtaining method, preprocessing is carried out on a multi-source heterogeneous data set, a standard data set is obtained, multi-modal feature extraction processing is carried out on the basis of the standard data set, and a multi-modal feature set is obtained; performing cross-modal alignment processing on the multi-modal feature set to generate a cross-modal alignment feature set, performing fusion processing based on the modal alignment feature set to generate a fusion feature set, and inputting the fusion feature set into the to-be-trained model to perform training to obtain a data processing model. According to the method provided by the invention, the adaptation capability of the data processing model to multi-source heterogeneous data and the expression capability of the data processing model to a high-dimensional structure are enhanced while the multi-modal semantic integrity is ensured, so that the semantic understanding and reasoning judgment performance of the model in a complex scene is remarkably improved.

The invention provides an intelligent text retrieval method and system, a storage medium and a program product, and relates to the field of intelligent text retrieval, the method comprises the following steps: segmenting each original document into a plurality of semantic segments according to semantic similarity; extracting a text feature of each semantic fragment to generate a text fragment vector; selecting a related fragment from each text fragment vector, wherein the first similarity between the related fragment and the retrieval vector is greater than a preset first threshold value; calculating a second similarity between each related fragment and the adjacent fragment; combining the adjacent segments with the second similarity greater than a preset second threshold value with the corresponding related segments to obtain stitching segments; based on a retrieval result corresponding to a historical retrieval request of a user, determining a retrieval granularity which is firstly displayed; and according to the retrieval granularity, adjusting the display content of the stitching fragment, and returning a text retrieval result. By implementing the method, the semantic integrity and the result relevancy of the text retrieval result can be improved.

The invention provides an agent visual language navigation method and system based on task completion prediction. The method comprises the step of constructing a dual-drive structure composed of a self-adaptive mixed pooling mechanism and a task completion analysis module. Firstly, in the visual information processing process, a dynamic weight distribution strategy is adopted to carry out multi-scale adaptive mixed pooling on panoramic features, so that the fusion effect of local and global semantic information is optimized, and the retention capability and semantic integrity of navigation historical information in a dynamic topological map are improved. And then, inspired by a human navigation cognitive behavior mechanism, a task completion analysis module is designed and introduced, and the task execution progress is dynamically estimated based on the recognition condition of a key landmark in a navigation path, so that an intelligent agent is driven to preferentially select a key path node and invalid exploration is reduced. And finally, realizing efficient understanding and execution of the natural language instruction by the intelligent agent through a multi-round cyclic cross-modal reasoning and action prediction mechanism.

The invention discloses a speech semantic analysis method and device, and a storage medium. The method comprises the steps of determining a real-time speech recognition result corresponding to frame-by-frame input speech; inputting the real-time voice recognition result and the context information into a semantic integrity judgment model to judge whether the real-time voice recognition result forms a complete semantic unit or not; and determining a real-time semantic analysis result corresponding to the real-time voice recognition result based on the streaming semantic analysis engine and the context information under the condition that the real-time voice recognition result is detected to form a complete semantic unit. Therefore, by introducing a frame-by-frame real-time speech recognition mechanism and a semantic integrity discrimination model, a processing link of staged dependence and waiting in a traditional speech processing system is broken, and closer dynamic cooperation between speech recognition and semantic analysis is realized.

The invention provides a mediation document generation method and system, relates to the technical field of data processing, and comprises the step of forming structured case data through text preprocessing, named entity recognition and legal information association based on case text data and legal corpus information. Then, generating a preliminary mediation document with placeholders by utilizing case classification, semantic matching and a sequence-to-sequence model; after standard auditing and logic auditing are carried out on the documents, optimization is carried out based on reinforcement learning, language style migration and a text generative adversarial network, and the structural rationality, law term normativity and semantic integrity are improved. And finally, a mediator preference template is constructed through Few-shot Learning and meta learning, the format and language style of the document are adjusted in a personalized manner, and a mediation document conforming to the habits of a mediator is generated. According to the method, the intelligence and the accuracy of mediation document generation are improved, the law compliance and the logic preciseness are ensured, and the mediation working efficiency is improved.

The invention discloses a cue word compression method and system based on TF-IDF and cross entropy, belongs to the technical field of large model cue word compression, and aims to solve the problems that redundant information is introduced into long cue words, the model efficiency is reduced and the cost is increased. To-be-compressed content is divided into sentences at the sentence level and then converted into embedded vectors, and the Euclidean distance is calculated in combination with problem vectors so as to screen related sentences; calculating a TF-IDF value at the word level through a word frequency and an inverse document frequency to extract keywords and recombine sentences; and selecting a reference model and a basic model at the Token level, identifying the key Token based on a cross entropy loss difference value, and splicing the key Token in sequence to generate a compressed cue word. According to the method, a complex calculation structure is avoided, the inference efficiency is improved while the semantic integrity is maintained, and the resource consumption is reduced.

The invention discloses a knowledge question-answering method and system based on topic knowledge graph retrieval enhancement, and the method comprises the steps: firstly extracting a local topic represented in a triple form based on an original document through employing a large language model, carrying out the clustering, and generating a global topic triple set representing the global perspective of the whole document; secondly, on the basis of the global topic triple set, topic-guided entity and relation extraction is adopted, and a mixed knowledge graph is constructed; secondly, providing a semantic perception personalized PageRank algorithm, matching query semantics with semantics of edges in the mixed knowledge graph, and dynamically adjusting the weight of score propagation between nodes; and finally, designing a three-level progressive retrieval mechanism, retrieving multi-level information related to user query from the mixed knowledge graph, and inputting the multi-level information into the large language model to generate a final answer. According to the method, the semantic integrity and retrieval precision of the knowledge graph are remarkably improved, and the accuracy, comprehensiveness and enabling performance of generated answers are ensured.

The invention discloses a text abstract generation method and system based on sparse attention acceleration, and the method comprises the steps: reading long text data, carrying out word segmentation and embedded coding processing, extracting a sequence feature vector, and mapping the sequence feature vector into a query matrix Q, a key matrix K and a value matrix V; constructing an abstract generation network, wherein the abstract generation network comprises a sparse attention calculation module, a feedforward calculation module, a prediction head module and a key value cache module; inputting the query matrix Q, the key matrix K and the value matrix V into an abstract generation network, passing through a backbone network formed by stacking a sparse attention calculation module and a feedforward calculation module for multiple times, processing by a prediction header module, and caching a historical decoding state in real time through a key value caching module to obtain an initial text feature vector; and based on the initial text feature vector, executing an autoregressive decoding process through an abstract generation network, and outputting a final abstract result. According to the method, the decoding process can be accelerated, and the semantic integrity and coherence of the generated abstract are ensured.

The invention discloses a recommendation system and method for keeping semantic integrity based on a large language model. According to the method, user-article interaction data and text information are fused, a prompt template of a user and an article is constructed, a configuration file with rich semantics is generated by utilizing a large language model, and initial embedded representation is extracted. Then, two-stage dimensionality reduction transformation is carried out through principal component analysis and a multi-layer perceptron, semantic information is reserved, and low-dimensional embedding is generated; on the basis of the embedding, cosine similarity is calculated, a user-user and article-article similar graph is constructed, and final embedding representation is generated through graph convolutional network coding. Meanwhile, collaborative filtering is combined to capture an interaction relationship and optimize a joint learning target, including recommendation loss, cross-modal comparison loss and regularization terms, so as to align semantics and collaborative filtering embedding, and finally generate a high-precision personalized recommendation result. The method effectively improves the semantic comprehension ability and recommendation accuracy of a recommendation system, and is suitable for various recommendation scenes.

The invention discloses a document segmentation method for large model retrieval enhancement generation, and relates to the technical field of document segmentation, the method comprises the following steps: taking any secondary title and text content under the secondary title as a to-be-analyzed object, and counting the number of characters in each to-be-analyzed object, when the number of the characters exceeds the preset number of the characters, paragraph splitting is carried out on the text content; and carrying out segmentation confidence calculation on each split part, and carrying out merging or segmentation processing on the adjacent parts based on the segmentation confidence corresponding to the adjacent parts. According to the method, the segmentation confidence coefficient of the adjacent parts is calculated, the document can be accurately and partially segmented, the segmentation effect is effectively guaranteed, and in addition, based on the segmentation confidence coefficient, the semantic integrity after segmentation and before and after segmentation can be effectively guaranteed.

The invention provides a task parallel processing intelligent scheduling method and system oriented to large model services, and belongs to the technical field of service computation.The method comprises the steps that semantic features of instructions and texts of parallel subtasks are extracted through a LaBSE model, embedded vectors are generated, the vector dimension of the embedded vectors is reduced through a compression module, and the embedded vectors are distributed in parallel; inputting the compressed embedded vector into a regression device to predict the generation length; constructing a segmentation algorithm based on semantic integrity constraint, and segmenting the text into a text fragment list; dividing the segmented text fragments in the text fragment list into different batches of tasks by adopting a multi-constraint task batching algorithm according to a similar length constraint and a same instruction constraint to obtain a batch task set; and optimizing the batch task set by adopting a task allocation algorithm oriented to heterogeneous resources, and determining an optimized sub-task scheduling scheme. According to the method, efficient parallel processing of large model services is realized, and the reasoning speed, the resource utilization rate and the intelligent level of task scheduling are remarkably improved.

The invention discloses a test case generation method, electronic equipment and a storage medium, and the method comprises the steps: carrying out semantic recursion slicing on a target document, and obtaining semantic coherent demand document slices; performing fragment sliding window processing on the document slices through a preset large language model to obtain a plurality of test cases; and according to the correlation of the functions contained in the plurality of test cases, merging the plurality of test cases to obtain a test case set. The invention discloses a test case generation method, electronic equipment and a storage medium, which can keep semantic integrity of a segmented text and correspond to context information, and realize that a demand document outputs a test case in an end-to-end manner.

The embodiment of the invention provides a retrieval enhancement generation method and system based on multivariate fusion, and the method comprises the steps: firstly structuring a knowledge base document, generating a knowledge graph, and importing an ES to establish an index; inputting question sentences, carrying out word segmentation and other processing, querying instance nodes from the map, and outputting meeting conditions according to answer sentence patterns; if not, fragmenting and blocking the document according to a title level, obtaining candidate results through vector, ES and atlas retrieval, normalizing scores, merging and optimizing the scores of the candidate retrieval results according to a retrieval source, calculating comprehensive scores, and outputting the comprehensive scores in a descending order; and finally, carrying out semantic integrity aggregation on the combined candidate results, and inputting into a large language model to obtain a final answer. According to the method, the relevance of the retrieved content is greatly improved, the semantic integrity of the retrieved content is improved, the model magic view problem is reduced, and the question and answer accuracy and the answer quality of a knowledge base are improved.

The invention discloses a long document knowledge base construction method based on directory perception and a program product, and belongs to the field of artificial intelligence and natural language processing. According to the scheme, an original long document is sequentially subjected to preprocessing, directory structure analysis, mixed blocking, double-tag generation, tag intelligent optimization, vectorization and meta-information mounting, and finally automatic construction and high-quality retrieval enhancement generation of a knowledge base are achieved. According to the method, the semantic integrity is guaranteed by fully utilizing the perceptual ability of the directory structure, the label quality and the retrieval efficiency are improved by combining a double-label system and intelligent optimization circulation, and the construction efficiency, the retrieval accuracy and the result traceability of the knowledge base are remarkably improved; the method is suitable for intelligent processing and application of long complex structure documents such as academic specialities, technical documents, policies and regulations and the like.

The invention discloses a global knowledge extraction method and system based on a large model and an RAG technology, and belongs to the technical field of large model data processing. According to the global knowledge extraction method and system based on the large model and the RAG technology, local sensitive hashing is adopted for duplicate removal, similar texts are combined through dynamic thresholds, and redundant data interference is reduced; a text segmentation strategy based on separator confidence ensures that segmented text blocks maintain semantic integrity and are adaptive to large model input length limitation, similarity retrieval and full-text retrieval are fused, results are reordered in combination with an RRF algorithm, semantic relevance and accuracy are considered, the multi-scene knowledge construction requirement is met, and the multi-scene knowledge construction efficiency is improved. High-correlation entity pairs are screened through mutual information, context knowledge is retrieved from a vector database in combination with an RAG system, semantic reasoning is conducted through a large language model, hidden logic relations among entities are generated, and jumping from data association to knowledge generation is achieved.

The invention relates to an intelligent document information real-time retrieval method and system based on the RAG technology, and the method comprises the steps: analyzing documents of various formats, extracting a text, maintaining the content continuity through adaptive semantic partitioning processing, and building a character-level position index at the same time; after vectorizing the text blocks, generating a plurality of rewriting queries for the original query; respectively carrying out mixed retrieval (combining keywords and semantic retrieval) for each rewriting query, and fusing and reordering results to obtain candidate text blocks; after correlation filtering, inputting a large language model according to correlation to generate an answer, and if the result is negative, triggering secondary retrieval and reordering; and finally, outputting a structured answer containing position information and supporting front-end visualization. According to the method, the semantic integrity maintenance, the multi-format document processing efficiency and the key information retrieval accuracy are effectively improved.

The invention discloses a large language model multi-source inference knowledge graph question answering method and system, and the method comprises the steps: extracting entities in a question, linking the entities with knowledge graph entities, and determining a subject entity set; searching paths in the knowledge graph by adopting a graph constraint reasoning method, and obtaining candidate answers and reasoning paths; detecting incomplete path coverage or answer dimension deficiency by using a large language model, triggering question decomposition to generate logically complementary sub-questions, and obtaining answers; establishing an inference source coordination layer, a coordination graph constraint inference source, a planning-retrieval inference source and a sub-problem inference source for multi-source evidence; and standardizing the output of each inference source into a structured evidence triple, sorting and selecting the first K evidences according to relevance, and generating a final answer by utilizing a large language model to carry out inductive inference. According to the method, reasoning can be dynamically adjusted, multi-source evidences can be effectively aggregated, logic consistency and semantic integrity are guaranteed, and the integrity and accuracy of knowledge graph multi-hop questions and answers are improved.

The embodiment of the invention provides a long text processing method and device, equipment and a medium. The method comprises the following steps: performing word segmentation on a long text to be processed to obtain a token sequence; the token sequence comprises a plurality of tokens; performing prediction processing on the token sequence by adopting a pre-acquired language processing model to obtain a processed token sequence; the processed token sequence comprises the token sequence and at least one prediction token behind the token sequence; the attention layer of the language processing model is obtained based on fusion of a state space model SSM and a self-attention mechanism, the SSM is used for obtaining global attention of the input token, and the self-attention mechanism is used for obtaining local attention of the input token. According to the method, when the long text is modeled, the effects of reducing the calculation complexity of the model and ensuring the semantic integrity and coherence of model text modeling are achieved.

The invention relates to the technical field of intelligent anti-fraud, in particular to a dynamic anti-fraud decision-making method based on adversarial evolution, which comprises the following steps: firstly, constructing an original fraud sample data set as the basis of adversarial sample generation and model training; the attacker model is based on a language generation model, generates a confrontation sample with sensitive word disguise, word order disturbance and semantic concealment features through guiding prompt, and simulates a real fraud variant. And mixing the confrontation sample and the original sample to construct a training set, inputting the training set into a defender model for training, and optimizing an attacker model prompt with recognition accuracy to realize dynamic iterative lifting. A dynamic purification and semantic compensation mechanism is introduced into the defender model, a gradient sensitive area is inhibited, semantic integrity is repaired, the stability and recognition capability of the model in a complex context are enhanced, and therefore accurate recognition and interception of diversified novel fraud texts are achieved.

The invention discloses a medical document slicing and retrieval method and device, equipment and a storage medium, and the method comprises the steps: generating a high-fidelity semantic summary text and a child node of a structured keyword set through a father node containing an original complete paragraph text, and storing the child node to a vector database; storing the semantic summary text and the structured keyword set into a keyword database; and when a user query instruction is received, performing dual-channel retrieval through the vector database and the keyword database, obtaining related target child nodes, extracting target original contents of a target father node from the target child nodes, transmitting the target original contents to the large language model, generating a final answer, and sending the final answer to the user. According to the method, the relevance and accuracy of retrieval results can be remarkably improved, instant and accurate decision support basis is provided for doctors, semantic integrity can be reserved, efficient multi-modal retrieval can be achieved, and the speed and efficiency of medical document slicing and retrieval are improved.

The invention provides a video multi-language conversion method and system based on semantic segmentation, which are used for solving the technical problems of semantic segmentation and low processing efficiency in the existing video multi-language conversion. The method comprises the following steps: firstly, performing audio-video separation preprocessing on an input video, and extracting an audio stream and a video stream; then voice feature extraction is carried out based on the audio stream, phoneme, pause and intonation features are obtained, and voice recognition is carried out to obtain text features; semantic segmentation points are determined according to the voice and text features, and semantic fragments are generated; the semantic fragments are converted into independent tasks, and multi-language conversion is executed through a parallel processing framework; and finally, synthesizing the processed semantic fragments to generate a video of a target language. According to the method, the semantic integrity is ensured through a semantic segmentation technology, the conversion efficiency is improved by adopting a parallel processing framework, and high-quality video multi-language conversion is realized.

The invention provides a sound wave signal self-supervised learning enhanced robust noise reduction processing method and system, and relates to the technical field of signal processing, and the method comprises the steps: carrying out the feature enhancement of an initial time-frequency representation through a dynamic adaptive mask strategy, constructing a self-supervised reconstruction task based on the mask time-frequency representation, separating noise and signal subspaces in a semantic manifold space, and carrying out the self-supervised learning enhanced robust noise reduction. And establishing a probability propagation network in combination with time domain continuity characteristics to model a local dependency relationship, and finally generating a noise reduction weight and realizing semantic fidelity optimization. The method can effectively improve the noise reduction effect and semantic integrity of sound wave signals in a noise complex environment.

The invention discloses a long text reading understanding method based on dynamic partitioning and selection, which is characterized in that the method adopts dynamic partitioning to dynamically divide long text input into discrete text blocks, selects and screens out irrelevant text segments by using selection partitioning, and splices the remaining text segments according to an original sequence, so as to obtain a long text reading understanding result. In order to conform to context window limitation predefined by a large language model, the method specifically comprises the steps of text preprocessing, dynamic blocking, block selection, large model output and the like. Compared with the prior art, the method has the advantages that the semantic coherence and the understanding accuracy are improved, the processing capability of the model on the super-long text is enhanced, the internal semantic integrity of each block is ensured, the semantic ambiguity caused by the block is reduced, the semantic coherence damage caused by the block with the fixed length is avoided, and the semantic coherence and the understanding accuracy are improved; and the processing capability of the model on the super-long text is enhanced, the data utilization efficiency is high, and the method has an important value and a good application prospect in practical application.

The invention discloses a multi-language text self-adaptive configuration method and electronic equipment, and relates to the technical field of computers, and the method comprises the following steps: obtaining a source text set in response to a translation instruction received when a page runs, and translating the source text set according to a target language identifier and a constraint strategy to obtain a translated text set; according to a constraint strategy, carrying out adaptability detection on each translation in the translation set to obtain an adaptive set and a non-adaptive set; performing semantic rewriting on each non-adaptive translation in the non-adaptive set to obtain a target candidate set; and aggregating the adaptation set and the target candidate set in the same rendering frame, and rendering and displaying the adaptation set and the target candidate set in batches, so that the problems of lack of multi-language dynamic adaptation capability and insufficient semantic equivalent compression are solved, the accumulated layout offset is remarkably reduced on the premise of ensuring semantic integrity and readability, and the layout efficiency is improved. And the page stability and the user experience are improved.

The invention provides a target-oriented video semantic communication system based on a visual model, and the system comprises a semantic extractor which is based on an SAM2 model and is used for processing an original video, generating a segmentation mask and extracting semantic information; the ViMama encoder is used for carrying out channel encoding on the output of the semantic extractor; the channel adaptation module is used for optimizing a coding sequence ViMama decoder based on the signal-to-noise ratio information of a physical channel, and is used for carrying out channel decoding to obtain a feature sequence; and the semantic reconstruction device is used for performing semantic reconstruction based on the feature sequence, recovering data and outputting a target video. According to the method, the problems of large redundant semantic information interference, insufficient deep semantic coding capability and poor communication robustness in a complex channel environment in a video are solved, efficient compression and robust transmission of video data are realized on the premise of ensuring semantic integrity, and the overall performance and adaptability of semantic communication are greatly improved.

The invention relates to the technical field of semantic processing and knowledge maps, and provides an AI-based intelligent question-answering method and system, and the method comprises the steps: collecting a knowledge map of an intelligent question-answering library, and obtaining an inquiry text of a user; performing word segmentation on the inquiry text to obtain a plurality of vocabularies; obtaining a reply design scale of the inquiry text; determining the number of analysis queues and establishing a plurality of analysis channels; obtaining the semantic intensity of each marked vocabulary and obtaining a high-intensity semantic vocabulary sequence; quantizing the analysis contribution of each mark vocabulary at each moment and the initial analysis limit at each moment, and further obtaining the buffer semantic integrity retention degree at each moment; judging and establishing a buffer space in real time, adjusting an analysis process, and finally outputting a semantic weight of each vocabulary; and screening a plurality of core vocabularies, obtaining a relation path in combination with the knowledge graph, and further generating a reply text. The invention aims to solve the problem that semantic weight acquisition is inaccurate due to word order interference in text keyword analysis.