2601 results about "Sentence" patented technology

A behavioral authorship verification system captures and analyzes multi-modal behavioral patterns during content creation to authenticate human authorship. The system comprises a processor executing behavioral analysis modules that generate comprehensive behavioral fingerprints distinguishing genuine human authors from AI-generated content and impostor authorship. A sentence progression mapping module detects sentence boundaries and captures intermediate composition states including additions, deletions, and modifications. A multi-modal input analysis module monitors keystroke dynamics including flight time and dwell time while detecting paste events and input method transitions. A behavioral pattern recognition engine generates user-specific baselines from historical sessions and computes deviation scores using statistical distance metrics. An anomaly correlation module aggregates behavioral deviation signals using weighted fusion algorithms to detect sophisticated mimicry attempts. An authorship scoring engine synthesizes outputs into unified confidence scores while maintaining temporal authorship chains. The system enables real-time authorship verification during content creation rather than post-hoc analysis.

The invention discloses a large-model real-time voice text intention recognition method and system based on text prediction, and the method comprises the steps: obtaining the real-time voice data of a user, carrying out the real-time voice recognition processing through a streaming voice recognition interface, and obtaining a part of transcriptional text; inputting the partial transcription text into a mask language model for text prediction, and generating a plurality of high-credibility complete sentence candidates; based on the complete sentence candidates, the complete sentence candidates are input into a large language model in parallel for intention recognition, a corresponding intention result is obtained, and a mapping relation between the candidate sentences and the intention recognition result is established; and obtaining a sentence completely expressed by the user, calculating the similarity between the complete actual sentence and a plurality of high-credibility complete sentence candidates through a multi-level text similarity algorithm, selecting the candidate sentence with the highest similarity score, and directly obtaining a corresponding final intention recognition result based on the mapping relationship. The objective of the invention is to solve the technical problem of high response delay of an existing voice intention recognition system.

A neural end-to-end system is provided for the face and voice preserving translation of videos. The system is a pipeline of multiple models that produces a video of the original speaker speaking in the target language with modified lip movement to match the target speech, while preserving emphases and prosody of the original speech, and voice characteristics of the original speaker. The pipeline starts with automatic speech recognition including emphasis detection, followed by the translation model. The translated text is then synthesized by a Text-to-Speech model that recreates the original emphases in the target sentence. The resulting synthetic speech is then converted back to the original speakers' voice using a voice conversion model. Finally, to synchronize the lips of the speaker with the translated audio, a generative model generates frames of adapted lip movements which are combined with the audio to produce the final output. The disclosure further describes several use-cases and configurations that apply these techniques to video conferencing, dubbing, low-bandwidth transmission, speech enhancement and assistive technology for the hearing impaired.

The invention discloses a video content semantic understanding and text description generation method based on deep learning, and relates to the technical field of multimedia information processing.The method comprises the steps that the semantic similarity of a text and a video frame is calculated through a CLIP model, related key frames are selected, and features are aggregated; respectively extracting audio, visual and semantic features; aligning different modal features by using self-attention, unifying dimensions of the LSTM, and then splicing and fusing; attention weights are calculated at a video level, a frame level and a channel level, and key information expression is enhanced; swin Transform encodes fusion features, and LSTM (Long Short Term Memory) decodes step by step to generate natural language description; and a text-video index database is constructed, and rapid retrieval is realized based on semantic similarity. According to the method, the mapping relation between the video features and the natural language is learned end to end through the deep learning model, dependence on a fixed template can be eliminated, and semantic description with various sentence patterns and coherent logic is generated.

The invention relates to the technical field of knowledge extraction, in particular to a knowledge graph construction method and system based on a large model, and the method comprises the following steps: obtaining a current input statement of a user through a dialogue state tracker, inputting the statement into a BERT intention classification model for domain label analysis, behavior type recognition and emotional tendency detection, and outputting a three-dimensional classification vector; and extracting entity lexical items and relation predicates based on an LSTM sequence tagging device, and generating an original semantic structural body. According to the method, intention classification, behavior recognition and emotion detection are fused through three-dimensional semantic analysis, semantic comprehension granularity is improved, dynamic entity disambiguation is combined with a Manhattan distance threshold value and dialogue history tracking, semantic boundaries are defined to reduce anaphora ambiguity, and cross-modal alignment is enhanced through relation predicate hierarchical clustering and knowledge base dynamic matching; generative reply and semantic coherence reordering collaboratively keep topic continuation, and structured analysis and unstructured generation closed loop optimize semantic output and interaction fluency.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using a set of large language models to determine a natural language response to a query. One of the methods includes receiving a query related to a document. The document is submitted to a first model along with a prompt to generate an outline of the document. The document is submitted to a second model along with a prompt to generate metadata of the document. At least a portion of the query, document metadata, and the document outline are submitted to a third model with a prompt to generate a natural language response to the query. A selected sentence from the natural language response is correlated to a document sentence. The natural language response is provided to the user with an indication that the selected sentence from the natural language response is correlated to the document sentence.

The invention relates to a document processing method and system based on text content extraction. The method comprises the steps that an original document containing text, image and format information is received, the encoding format of the document is automatically detected, character set conversion is executed, and hierarchical indexes including page numbers, paragraphs and tables are established for an unstructured document; the method comprises the following steps: synchronously processing text content and visual layout through a pre-trained visual-language model, extracting word-level and sentence-level semantic features by a text stream embedding layer, analyzing spatial distribution features of document elements by a visual encoder, and fusing text and visual features through a cross-modal attention mechanism; and loading the domain knowledge graph matched with the document type, and executing entity linking to associate the text mentions to the knowledge nodes. According to the document processing method and system based on text content extraction, through the synergistic effect of vision-text joint coding and knowledge enhancement, the accuracy of financial contract key clause recognition tasks is improved, the error rate is lower than that of industry benchmark products, and the semantic understanding precision is remarkably 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 discloses an intelligent report generation method and system based on multi-source heterogeneous data fusion, and particularly relates to the technical field of data fusion, and the method comprises the steps: constructing a fact anchor point containing a stable identifier in a fusion layer, and solidifying an aperture version signature and a time interval; packaging a screening condition, a sorting rule, an access plan, time window limitation and the like into a controlled context packet and freezing the controlled context packet into a unique fact source; in the range, generating sentences by using a limited template and writing sentence-level reference marks, so that each sentence can be traced back to a data slice and an evidence fragment; performing consistency verification according to a source, an aperture and time, performing conservative degradation on the attribution expression according to a consistency threshold value, and outputting a degradation strategy description; and finally, page two-way pointer and cross-sentence consistency recording and audit playback are realized through an evidence chain pointer table and a published version signature, and the problems of caliber drift, cross-window access and difficulty in evidence tracing are solved.

The invention discloses a text retrieval method and device, equipment, a storage medium and a product. The method comprises the steps of obtaining a question input by a user and at least one candidate text; for each candidate text, determining semantic similarity and keyword similarity between the question and the candidate text; sentence features of the questions and document features of the candidate texts are extracted respectively, and the weight of semantic similarity and the weight of keyword similarity are determined according to the sentence features and the document features; based on the weight of the semantic similarity and the weight of the keyword similarity, according to the semantic similarity and the keyword similarity of the question and each candidate text, calculating the comprehensive similarity of the question and each candidate text; at least one candidate text with the maximum comprehensive similarity is screened out, and an answer candidate set is obtained; and inputting the answer candidate set and the question into a preset large language model, and outputting a retrieval result. According to the method, the accuracy and efficiency of long text retrieval can be improved.

The invention discloses a database question and answer model training method and device, a storage medium and computer equipment, and the method comprises the steps: associating standard structured query language statements, standard execution result answers and standard natural language questions, and generating training annotation data; and collecting simulation derivation problems possibly proposed for the database to obtain non-labeled data for training. Based on a GRPO reinforcement learning framework and a scoring reward function provided by a double-tower model, training is carried out on the scoring reward function by utilizing training labeling data, supervised fine tuning training is carried out on a database question and answer model, and non-labeling data for training, format rewards, executable rewards and scoring rewards of the scoring reward function are combined, so that the scoring reward function of the database question and answer model is obtained. And continuing to train the database question and answer model after supervised fine tuning training. Preliminary training is carried out through a small amount of annotation data, then subsequent training is carried out through non-annotation data, the reasoning ability of the model can be stimulated, the annotation cost is reduced, and the training efficiency is improved.

The invention belongs to the technical field of text processing, and particularly relates to a method and system for retrieving DOCX document content based on keywords, which comprises the following steps: analyzing an Office Open XML structure of a DOCX document, combining with multi-dimensional features such as style names, and utilizing a title classification score model to accurately distinguish a title and a text, so that a semantic hierarchical structure of the document is effectively reserved; and secondly, a multi-level semantic extension mechanism is introduced, and a Sension-BERT, a HowNet knowledge base and a Word2Vec model are fused, so that intelligent extension of synonyms and synonyms of keywords is realized, and the recall rate and semantic understanding ability of retrieval are remarkably improved. And in addition, a BM25 model is combined with paragraph length normalization and structure position weight to calculate a correlation score, so that retrieval results are sorted more accurately and reasonably. The construction of the reverse index is combined with the position coding and compression optimization strategy, and the retrieval efficiency and the storage performance are both considered.

The invention relates to the technical field of natural language processing, in particular to a context-based Prompt generative model optimization system, which comprises a context analysis module, a cue word generation module, a context optimization module, a semantic check module and a structure reconstruction module. According to the method, the context path and the semantic hierarchy information of the semantic unit are introduced, the fine degree of semantic matching degree recognition is improved, semantic guide deviation caused by statement template solidification is avoided, the cue words are recombined in combination with the semantic coherence weight and the logic dependency relationship, and the recognition accuracy is improved. The consistency and expression accuracy of the prompt content in the context are enhanced, the prompt word insertion sequence and connection mode are dynamically adjusted through a semantic conflict detection and structure rechecking mechanism, coherence and stability of a semantic structure and controllable generation of the prompt content are kept, semantic conflicts and expression chaos caused by static matching are effectively avoided in the generation process, and the generation efficiency is improved. And dynamic adaptation of prompt configuration and smooth optimization of language output are integrally realized.

The invention discloses a large language model training method and device, electronic equipment and a storage medium, and relates to the technical field of artificial intelligence and reinforcement learning. The method comprises the steps of obtaining an initial prediction answer statement corresponding to a question statement; semantic expansion retrieval is carried out on the initial prediction answer statement, and a retrieval answer statement corresponding to the initial prediction answer statement is determined; scoring the retrieval answer statement, and determining a target reward value of the initial prediction answer statement based on a scoring result of the retrieval answer statement; and if the target reward value is smaller than a preset reward threshold value, adjusting the pre-trained large language model based on the target reward value, and inputting the question statement into the adjusted model again until the finally obtained target reward value is greater than or equal to the preset reward threshold value, thereby completing training of the pre-trained large language model. The optimization training efficiency of the large language model and the accuracy of model output can be improved, and the stability of large language model output is improved.

The embodiment of the invention provides a multi-modal fusion entity retrieval enhancement generation method and device, and the method comprises the steps: carrying out the blocking and adaptive text extraction of multi-modal data in an offline stage, obtaining the text block data corresponding to each modal data, extracting the entity and relation of the text block data according to a language model, constructing an entity triple, and carrying out the segmentation and adaptive text extraction of the entity triple. Fusing the entity triad with the text block data to obtain an offline knowledge graph, and constructing a data index; in the present stage, a query statement of a user is received, text block data most similar to the query statement are retrieved in a knowledge graph through a data index, after entity aggregation is carried out on the retrieved text block data, the text block data are reordered according to retrieval scores, an entity aggregation result is obtained, entity ordering is carried out according to the entity aggregation result, and the entity aggregation result is obtained. By means of the multi-modal retrieval method and device, the efficiency and accuracy of multi-modal retrieval can be improved.

The invention discloses a text sentiment analysis method, system and equipment based on multi-granularity sentiment modeling and a medium, and the text sentiment analysis method comprises the following steps: obtaining a to-be-analyzed initial text, and carrying out standardized preprocessing on the initial text to obtain text data; performing coarse-grained sentiment analysis on the text data by using a chapter-level encoder to generate chapter-level sentiment tags; performing fine-grained sentiment analysis on the text data by using a sentence-level encoder to generate a sentence-level sentiment tag; performing local correction on the sentence-level emotion label based on the chapter-level emotion label by using a cross-layer attention mechanism to obtain an updated sentence-level emotion label; extracting entity features and attribute tags in the text data, and associating the entity features, the attribute tags and the updated sentence-level emotion tags to generate an emotion triple; and carrying out conflict analysis on the emotion triad to obtain a structured emotion label of the initial text. According to the invention, the context consistency and accuracy of the sentiment analysis result can be improved.

The invention relates to the technical field of English language ability assessment, in particular to an English language ability assessment method and system. The method comprises the following steps: collecting a writing text uploaded by a student, and analyzing the literature and style category and sentence pattern structure of the writing text to obtain a content literature and style category sentence pattern structure; thirdly, identifying logic relation deviation in the text based on the structures, performing clustering, further quantizing the context fuzzy degree, and generating context fuzzy degree quantized data; and finally, designing an English language ability evaluation architecture by adopting a Q-learning algorithm. According to the invention, the English language ability evaluation technology is optimized, so that the English language ability evaluation technology is more accurate.

The invention provides a large language model illusion suppression method, product and equipment based on multi-order detection. The method comprises the following steps: step 1, extracting and enhancing a core entity; 2, generating an initial answer; step 3, sentence detection and entity detection; 4, illusion is corrected; and step 5, performing iterative optimization. According to the method, the credibility and rationality of the generated text are improved.

The invention relates to a construction method of a letter intelligent question-answering system based on self-adaptive knowledge graph enhanced LLM, and the method comprises the steps: obtaining the text data of a target region, carrying out the preprocessing of the text data, converting the text data into a vector format through a fine-tuning sentence converter, and constructing a vector database; vectorizing the user question text and the identity information, calculating the similarity of the user question text and the identity information with a vector database to generate context information, jointly inputting the context information and the user question into a generative large language model, and outputting a plurality of candidate answers; and based on a pre-constructed target domain knowledge graph, carrying out joint reasoning verification on the plurality of candidate answers, sorting the candidate answers through a scoring mechanism, and selecting a final answer. Complex questions can be answered, and the accuracy and reliability of questions and answers are improved; the answer can be accurately generated through reasoning without based on a preset answer template, and the accuracy and logicality of the answer are ensured.

The invention discloses a natural language query method and device in the electric power field, terminal equipment and a storage medium, and belongs to the field of natural language process.The method comprises the steps that a natural language query statement input by a user is received, semantic disambiguation is conducted on the natural language query statement according to the historical context of the natural language query statement, and the semantic disambiguation result is obtained; obtaining a target entity; according to the natural language query statement, the target entity and a preset task target, constructing a cue word, and inputting the cue word into a pre-trained large language model to obtain a first SQL statement; executing the first SQL statement to obtain structured service data, and constructing a first cause and effect graph according to the structured service data; and optimizing the first SQL statement according to the cause and effect graph to obtain a second SQL statement, and executing the second SQL statement to obtain a query result. The problem of low power data query accuracy in the power field can be solved.

A method for updating a tabular search large language model (LLM) includes performing data pre-processing on new data associated with the tabular dataset to obtain a set of sequences, applying a first fine-tuning operation on the tabular search LLM using the set of sequences, applying a second fine-tuning operation on the tabular search LLM using training data to obtain a set of final loss results and a set of updatable gradients, wherein the training data comprises at least the set of sentence predictions, applying an optimization function on the set of final loss results and the set of updatable gradients to obtain optimized gradient descent parameters, and applying the updated tabular search LLM to a new input associated with the new data to produce a new output.

The invention discloses an intelligent translation calibration method based on an RAG knowledge base, and belongs to the technical field of artificial intelligence and natural language processing. In order to solve the problems that an existing large language model (LLM) is insufficient in professional term translation accuracy in text translation, low in RAG system retrieval efficiency and the like, an intelligent calibration triggering mechanism is constructed, so that the LLM can autonomously judge whether translation content needs knowledge base enhancement or not in the modes of field detection, term density analysis, user-defined triggering and the like; a full-amount retrieval strategy is changed, and the system efficiency is improved; an RAG knowledge base retrieval framework based on function calling is constructed, multiple types of knowledge sources such as a translation memory library, a term table and a professional corpus are integrated, and accurate domain knowledge is provided for translation calibration; flexible multi-knowledge-base configuration and priority management are supported, and a user can select knowledge base types and set retrieval strategies and fusion weights according to needs; through a progressive translation optimization process of initial translation, intelligent judgment, knowledge base retrieval and calibration optimization, accurate calibration of different granularities such as term level, sentence level and paragraph level is realized, and a standardized knowledge base interface specification is provided. According to the method, the defects of LLM in the aspects of translation accuracy and knowledge base utilization efficiency in the professional field are effectively overcome, and the accuracy and efficiency of knowledge-intensive translation tasks are remarkably improved.

The invention discloses a fusion semantic vector space mapping-based translation ambiguity term accurate matching method, which comprises the following steps of: S1, obtaining source language ambiguity terms, context texts and a target language candidate translation list, and extracting domain tags and term matching features to form a multi-modal data set; s2, using improved XLM-R model coding to generate term-level, sentence-level and translation-level semantic vectors; s3, training a dynamic mapping matrix based on a bilingual parallel corpus, and aligning source side vectors to a shared semantic space; s4, fusing the source-side basic vector and the multi-dimensional features through a double-channel attention fusion network, and generating source-side and translation-side comprehensive semantic vectors; s5, introducing term-context attention weight to correct cosine similarity; and S6, outputting an optimal translation through normalized sorting and part-of-speech secondary judgment. According to the method, multi-field ambiguous term accurate matching is realized, the term translation precision and efficiency in professional fields are improved, and the requirements of high reliability of term translation in the fields of medicine, machinery, computers and the like are met.

The invention relates to the technical field of electric digital data processing, in particular to a data acquisition method and device based on a large language model and a medium. The method comprises the steps of inputting a target statement into a large language model to obtain a target text output by the large language model, and obtaining a target keyword set of the target statement according to the target text; obtaining a target website list matched with the target keyword set from a preset website library; determining a target crawling time period of the website corresponding to each website and a crawling time step length corresponding to the target crawling time period according to a historical updating moment set of the website corresponding to each website in the target website list; and in the target crawling time period of the specified website, judging whether the data of the specified website is updated or not by taking the crawling time step length corresponding to the target crawling time period of the specified website as a judgment period, and if so, crawling the data from the specified website. According to the method, the data related to the statement input by the user can be comprehensively and effectively collected.

The invention provides an enterprise knowledge graph automatic construction and intelligent retrieval method, which comprises the following steps: collecting multi-source data from a heterogeneous enterprise information system, and carrying out data cleaning and standardization processing; on the basis of a comprehensive scoring mechanism of field similarity and behavior semantic vectors, entities from different systems are merged, and a standard entity set with a unique identifier is generated; based on the standard entity set, in combination with a scoring mechanism of a task-type relationship and a collaborative relationship, extracting a semantic relationship from a behavior record, and constructing an enterprise knowledge graph structure; extracting representative semantic paths from the knowledge graph structure, screening high-quality paths through a path scoring model, and organizing the high-quality paths into a structured path index set; and receiving a natural language query statement, encoding the natural language query statement into a semantic vector, matching the semantic vector with the path index set, executing query in the atlas in combination with an authority control mechanism, and returning a result.

The invention relates to the technical field of large language models, in particular to an official document content generation and compliance verification system based on a large language model. Comprising an official document extraction module used for analyzing a model essay by using a large language model, extracting a writing style and a fixed sentence pattern, constructing a knowledge graph, storing format specifications, common expressions and logic structures, mining and constructing a core intention library, and storing typical intention prototype vectors; the content generation module is used for generating paragraphs through context sensing, expanding and writing keywords, analyzing behavior data, predicting writing intention vectors and performing condition control; the compliance verification module is used for warning improper words in real time, retrieving policy verification logic and verification formats, and performing pre-verification based on intention vectors; and the writing auxiliary module is used for providing real-time modification suggestions and forward-looking prompts based on writing intention vectors. According to the technical scheme, document writing efficiency and quality can be improved.

The invention discloses a natural statement decoding method and device based on high-density electrocorticogram. The method comprises the following steps: acquiring an electroencephalogram signal acquired based on the high-density electrocorticogram; different frequency band signals are extracted from the electroencephalogram signals; the signals of different frequency bands comprise high gamma frequency band signals and at least one frequency band signal with the frequency lower than that of the high gamma frequency band signals; acquiring a voice starting point corresponding to each frequency band signal, and determining a target voice starting point based on the voice starting point corresponding to each frequency band signal; after the target voice starting point, acquiring a syllable classification result and a tone decoding result corresponding to each frequency band signal, determining a target syllable classification result based on the syllable classification result corresponding to each frequency band signal, and determining a target tone decoding result based on the tone decoding result corresponding to each frequency band signal; and determining a target tone language corresponding to the electroencephalogram signal based on the target syllable classification result and the target tone decoding result. According to the scheme, the natural statement decoding accuracy can be improved.

The invention relates to the technical field of intelligent question answering, in particular to a knowledge tracking and evidence labeling combined RAG question answering system which comprises a semantic analysis module, a statement positioning module, a knowledge path calibration module, an evidence index module and an enhanced question answering module. According to the method, through structural analysis of natural language input and extraction of clear verb semantics, index entities and time normalization expression, it is ensured that input statements have uniformity and processing consistency in the structural level, through a text segment screening strategy based on labels, the text scanning range is effectively limited, the index positioning precision and efficiency are improved, and the method is suitable for popularization and application. A causal node link with statistical significance in a text is clearly extracted, and index combination constraint and identification labeling are performed on key sentences in path nodes, so that the index cross coverage of candidate evidence fragments is enhanced, and the density of index integration in answers and the consistency of a logic sequence are improved.

A computer-implemented method for performing searches in a document database is disclosed. The method comprises automatically detecting a line of business associated with a user, receiving a text query from the user, and generating a query embedding from the text query. The method further comprises scoring entries in a reverse index using a hybrid scoring function. The reverse index comprises titles, title embeddings, sentences, sentence embeddings, and entity tags corresponding to documents in the document database. The hybrid scoring function is used to generate a score based both on a keyword match score between the text query and the reverse index and on a cosine similarity score calculated from embeddings in the text query and in the reverse index. The method also comprises ranking scores for sentences in the document database, and displaying a sentence associated with a top score to the user.

The invention discloses a speech training method and system based on artificial intelligence and application, and relates to the field of speech training, and the method comprises the steps: obtaining multi-modal data in real time when a user reads each training statement in each speech training process; obtaining a real-time multi-dimensional speech evaluation result according to the multi-modal data, and judging whether a next training statement exists in a speech training scheme in each speech training process or not; if not, generating a multi-dimensional speech evaluation result of each speech training process; if yes, the difficulty of the next training statement in the speech training scheme is adaptively and dynamically adjusted according to the real-time multi-dimensional speech evaluation result; and outputting the next training statement according to the adjusted speech training scheme, and repeating the steps until the next training statement does not exist in each speech training. The speech training scheme can be adaptively adjusted in real time in the speech training process, and the user experience, the participation degree and the training effect are remarkably improved.