2002 results about "Syntax" patented technology

The invention discloses a method for constructing a knowledge graph based on a large language model and a vector library, and relates to the technical field of knowledge graphs, and the method comprises the following steps: carrying out word segmentation, part-of-speech tagging and semantic analysis processing on a collected natural language text by using a pre-trained RoBERTa model to obtain a semantic representation vector of the natural language text; through a Seq2Seq framework and an attention mechanism, obtaining association between the semantic representation vector and the corresponding named entity, and obtaining the named entity through the obtained association; constructing a knowledge graph; fusing the constructed knowledge graph with a relational database of a preset field to generate a fused knowledge graph; named entities and corresponding attributes are obtained through dependency syntax analysis, the entity relation of the named entities is obtained, and a structured triple fusing the named entities, the attributes and the entity relation of the knowledge graph is constructed. In order to solve the problem of high entity linking difficulty in the prior art, the knowledge graph is converted into low-dimensional dense vector representation, and an inverted index database and the like are constructed, so that the difficulty is reduced.

The invention discloses a self-adaptive software vulnerability repair system based on deep learning. The self-adaptive software vulnerability repair system comprises a multi-modal feature fusion engine, a spatio-temporal joint modeling module, a vulnerability repair strategy generator and a cross-language representation model. The multi-modal feature fusion engine generates a mixed feature vector of a code, inputs the mixed feature vector into the spatio-temporal joint modeling module, captures code structure and time sequence dependence, and identifies risk features; the vulnerability repair strategy generator generates a repair scheme based on the risk features and optimizes the repair scheme through static verification; and the cross-language representation module performs feedback training by using the newly added sample and outputs a final detection result and a repair scheme. According to the method, the vulnerability trigger points can be accurately positioned, all the associated code snippets are extracted, a solid foundation is provided for subsequent vulnerability analysis, and the repaired codes conform to grammatical rules and keep original semantic consistency.

Implementations of this disclosure provide a machine learning model training system that receives user input being a natural language description of a search query, and packages and transmits the natural language description as a prompt to a plurality of large learning models (LLMs). The model training system also receives response from the plurality of LLMs being translations of the natural language descriptions to an executable search query and displays the translations to a user via a graphical user interface. The model training system receives user feedback via the graphical user interface that corresponds to indications as to whether each translation is correct, syntactically and / or semantically, and, in some examples, an indication of which response was preferred. The model training system also generates training data from the user input, translations generated by the plurality of LLMs, and user feedback, and subsequently, initiates training of a LLM using the training data.

The invention discloses a data query method and system based on a natural language text. The method comprises the following steps: dynamically shielding the natural language text to generate a semantic vector sequence; key entities and association relationships thereof in the semantic vector sequence are extracted, and a structured entity relationship set is obtained; matching the structured entity relationship set with knowledge graph nodes to generate a heterogeneous semantic graph; performing heterogeneous fusion on the heterogeneous semantic graph and text syntactic features to generate intermediate query representation; processing the intermediate query representation to generate an executable statement carrying a semantic association identifier; and verifying the semantic consistency of the response content of the executable statement and the original input, and iterating to regenerate the executable statement when the semantic consistency does not pass the verification. And efficient and accurate natural language query conversion and response generation are realized through dynamic shielding enhanced semantic robustness, knowledge graph path extension and weight iterative optimization, heterogeneous feature fusion and a multi-modal verification mechanism.

The invention discloses a CAPEC vulnerability management system based on a large language model, and belongs to the technical field of network security. The system comprises three modules: a vulnerability-CAPEC dynamic mapping module jointly encodes vulnerability description and code context through a bimodal large language model, accurately associates vulnerability logic with a CAPEC attack mode in combination with comparative learning and knowledge graph construction, and breaks through semantic limitation of traditional rule matching; the adversarial repair code generation module is used for generating high-robustness repair codes through adversarial training and syntax tree verification by fusing CAPEC relieving suggestions and code features on the basis of the association result, so that the secondary vulnerability risk is remarkably reduced; and the full-process automatic verification and DevOps integration module performs multi-dimensional security verification such as symbolic execution, fuzzy testing and the like on the generated repair code, and deeply integrates a development tool chain to realize real-time pushing and closed-loop management of a repair scheme.

The invention relates to a knowledge-enhanced large model data analysis agent visualization method, which comprises the following steps of: performing syntactic analysis by using a PCFG algorithm, identifying syntactic components and relationships of sentences, and splitting the sentences into sub-sentences with complete logic; extracting keywords and entity words of the sentences by combining a word segmentation technology with a business word segmentation library; recommending and summarizing candidate segmented words through a recall strategy, carrying out rough arrangement, fine arrangement and fine arrangement on the candidate segmented words, and correcting and complementing a segmented word result; splitting and analyzing the natural language problem based on the fine-tuned large model, and generating SQL statements meeting business requirements; analyzing the data set generated by the SQL, extracting field information and business meanings, and identifying an association relationship between fields; and automatically recommending chart styles according to data features and user preferences, assembling an analysis report and distributing the analysis report according to a specified frequency. According to the method, knowledge enhancement and large model technologies are combined, so that the efficiency and effect of data analysis and visualization are remarkably improved.

The invention relates to an intelligent semantic analysis knowledge graph system for multi-source heterogeneous data. The system comprises a data preprocessing module used for preprocessing to-be-processed multi-source heterogeneous data to obtain format-standardized text data; the semantic analysis module is used for performing natural language processing on the text data to obtain words, and then performing vocabulary layer, syntactic layer and semantic layer analysis on the text data and the words to obtain structured data; the structured data comprises a triple consisting of entities, relationships and events and a network diagram of the entities and the relationships; the model structure mapping module is used for mapping the structured data into a predefined structure of a knowledge graph to obtain structured graph data; and the knowledge graph module analyzes entities and relationships in the graph data, performs reasoning to obtain potential relationships in the graph data by utilizing a logic rule and a machine learning algorithm, realizes cross-data-source and cross-document deep semantic association, and improves the efficiency of large-scale data processing and knowledge management.

The invention provides an AI-based virtual digital human interaction system, and relates to the technical field of data processing, and the system is used for obtaining a user feature data set by performing feature extraction on an interaction data set, calculating an audio change rate and speech speed stability, obtaining speech emotion data, calculating syntactic complexity and counting the number of emotion words. Calculating a facial expression symmetry degree; obtaining facial emotion data; calculating body stability and identifying a visual staying area; obtaining attention state data; adjusting tone and speed of the virtual digital person according to the virtual digital person text, optimizing sentence pattern complexity and sentiment words in the virtual digital person text, adjusting changes of eyes, eyebrows and mouth corners of the virtual digital person, and adjusting information display sequence and visual focus guidance. According to the method, the interaction efficiency of the virtual digital human is improved by analyzing the facial expressions and the body actions of the user.

The invention discloses a cue word compression method and device of a large language model and a medium. The method comprises the steps that redundant vocabularies and redundant statements are recognized and removed according to the size relation between the vocabulary importance score and the importance threshold value of each vocabulary after original cue word segmentation; constructing a local syntax tree of each statement, and constructing a global analysis tree by the local syntax trees according to the dependency relationship among the cue words; calculating a local information entropy so as to align the LLM word segmentation device with nodes on the global analysis tree and adjust node values of the global analysis tree; trimming the global analysis tree through a recursive algorithm according to the adjusted node value, and generating a compression prompt corresponding to the prompt word; according to the method, the compression prompt is evaluated, the compression prompt is iteratively optimized according to an evaluation result, the cue word is reconstructed according to the optimized compression prompt, and compression of the cue word is achieved.

The invention relates to a sentiment analysis method based on multi-source knowledge and multi-granularity image-text features, which comprises the following steps of: A, acquiring text and picture data, identifying and marking aspect words and sentiment polarity thereof in the text, and forming a training set; step B, constructing a deep learning network model based on external multi-source knowledge and multi-granularity image-text features, obtaining knowledge-enhanced text feature representation and image feature representation by using an AMR image and an image tag, and obtaining a knowledge-enhanced text feature representation and a knowledge-enhanced image feature representation by using a syntactic dependency relationship of a text, a component tree structure of the text and a fine-granularity image-text relationship matrix; a multi-granularity text-visual fusion feature representation is obtained in combination with a multi-layer graph attention network; using the training set to train a deep learning network model; and C, sequentially inputting the text data and the picture data into the trained deep learning network model, extracting aspect words in the text data, and predicting the emotion polarity corresponding to the aspect words. The method and the system are beneficial to improving the accuracy of emotion polarity analysis.

The invention discloses a variation test method and device based on an intelligent automation script. The method comprises the following steps: extracting a statement execution sequence and a branch dependency relationship in an original source code by using a context-aware syntax tree, and generating a control flow diagram and a data flow diagram; performing code semantic analysis on the multi-dimensional code features by using a large language model, identifying potential defect types and variation rules, predicting the test efficiency of variants in combination with a deep Q learning model, and generating to-be-processed variants; performing grammar check and equivalence analysis on the to-be-processed variants to obtain effective variants, and testing the effective variants; variation test indexes are calculated, code weak points are positioned, and test blind areas are identified according to survival variants, so that a visual analysis report is generated; and iteratively generating a new test case according to the current survival variant and the visual analysis report, and testing the current survival variant by using the new test case. According to the invention, the variation test efficiency can be improved.

The invention discloses an attention aggregation and cross-modal graph fused unsupervised multi-modal emotion recognition method. The method specifically comprises the steps of 1, acquiring audio modal features, image modal features and text modal features of a video sample; step 2, acquiring an audio vector, an image vector and a text vector corresponding to the audio modal feature, the image modal feature and the text modal feature; step 3, constructing an unsupervised multi-modal emotion recognition model; 4, inputting the audio vector, the image vector and the text vector into an unsupervised multi-modal emotion recognition model for training, and obtaining the optimal output of the model as a final feature vector for prediction and classification; and 5, performing emotion recognition by using the trained unsupervised multi-modal emotion recognition model. The problem that in the prior art, the long-distance spatial dependency relationship of a tree or graph network cannot be processed, the syntactic dependency relationship in sentences cannot be captured, and consequently the model cannot make full use of multi-modal data is solved.

The invention discloses a structured text code generation and review method, which comprises the following steps of: receiving a control task demand input by a user, supporting natural language, structured form or graphical process input, performing deep semantic analysis by utilizing a demand analysis module, identifying control conditions, control actions and constraint factors, and generating and reviewing a structured text code. The task is converted into a plurality of control units through the SCoT chain; the SCoT chain reader generates a semantic vector according to a task demand, the AST reader generates AST representation through a grammar rule structure and a hierarchical relationship, and the decoder predicts a next grammar rule and generates an ST code in combination with the SCoT chain and AST features; and verifying grammar and logic of the ST code through an interactive automatic review and repair mechanism, if compiling errors occur, analyzing errors by the teacher model and providing repair suggestions, and performing code repair by the student model according to guidance until the code conforms to an expected function and passes compiling.

A context-based cyberattack signature generation system (“signature generation system”) disclosed herein comprises a signature prompt schema for generating prompts to a language model that generates cyberattack signatures. The signature prompt schema comprises a description of syntax for the cyberattack signatures and descriptions of case knowledge and domain knowledge for a type of cyberattack corresponding to a cyberattack signature. The signature generation system tests cyberattack signatures generated with the signature prompt schema against minimum signature conditions and traffic with ground-truth malicious / benign labels. Once the signature prompt schema passes the tests, the signature generation system deploys the tested signature prompt schema in combination with the language model for cyberattack signature generation.

The embodiment of the invention provides a code exception analysis method and device, a storage medium and a program product. In the embodiment of the invention, in the preset plug-in, the global knowledge graph is constructed through the logic syntax tree of the plurality of original business codes, so that intelligent anomaly analysis is carried out on the modified to-be-detected codes, multi-language and heterogeneous programming styles and complex dependency limitation are broken through, and cross-language and cross-file analysis is supported; change code snippets and associated influence code snippets thereof are accurately identified, and a data basis is tamped and analyzed; the analysis depth is deepened and the accuracy is improved through a multi-dimensional anomaly analysis large model and four-dimensional collaborative analysis in combination with an adaptive review rule and a knowledge graph; and finally, performing correction based on the generated repair scheme. Through the technical fusion of an MCP protocol, a multi-AI model, a syntax tree and a knowledge graph, the high efficiency, accuracy and real-time performance of code examination and repair are realized, and the service development quality and the system operation stability are remarkably improved.

The invention discloses a natural language-to-SQL (Structured Query Language) method and system based on RAG (Random Access Gateway) enhancement, and the method comprises the steps: constructing a query intention graph through dependency syntax analysis, recognizing and complementing semantic missing components, and forming complete semantic representation; vector representation is carried out on the business term segments by adopting vectorization coding, accurate definitions of business terms are obtained from a factory structured knowledge base, and an enhanced context set is formed through expansion retrieval in a low-confidence region; an SQL template mapping network is established based on historical query records, a mapping relation matrix is generated through field candidate matching, mapping conflict positions are identified, and multiple SQL candidate sequences are generated; grammar verification is carried out on the candidate sequence to identify grammar errors, semantic consistency verification is carried out to calculate the intention alignment degree, and an optimal SQL statement is selected through a deviation correction factor; and performing formatting processing and statistical abstract on an execution result, and providing data query capability for scenes such as factory quality management and equipment maintenance.

Systems, methods, and devices that relate to intelligent query decomposition and parallel routing for specialized model processing are disclosed. In one example aspect, the system receives a query from a user comprising a request relating to a particular domain. The system determines, using a decomposition model, a set of sub-queries based on semantic boundaries, syntactics, tasks, relationships, and rules relating to particular domains. The system inputs the set of sub-queries into a routing model to determine a set of specialized models. For each sub-query, the system routes the sub-query to a respective specialized model, generates an output, and assigns a confidence score. The system detects conflicts among outputs using a conflict detection model configured to identify discrepancies. The system generates an aggregated output by combining outputs according to a weighted aggregation algorithm prioritizing higher confidence scores and conflict resolution rules, then displays the aggregated output.

The invention discloses a Tibetan language multi-dialect real-time semantic conversion method based on a cross-language BERT model, and the method comprises the following steps: S1, collecting original text corpora of each dialect of the Tibetan language, and constructing a standardized training corpus set; s2, performing parameter initialization on the mBERT model, and preliminarily training the mBERT model; s3, constructing a semantic modeling model, and performing fine adjustment on the semantic modeling model; s4, receiving to-be-converted text input, and encoding; s5, obtaining an intermediate semantic representation vector of the text through a semantic coding sub-module; s6, inputting the intermediate semantic representation vector into a semantic generation sub-module, and generating target text output; and S7, executing syntactic consistency correction and language fluency correction. According to the method, mBERT modeling and an adversarial optimization mechanism are fused, real-time semantic consistency conversion of multiple dialects of the Tibetan language is achieved, and the method has the advantages of being high in accuracy, high in robustness and low in response delay.

The invention relates to the field of natural language processing technology, computer vision and multi-modal sentiment analysis, in particular to a multi-modal aspect-level sentiment analysis method based on multi-scale text visual feature enhancement, which comprises the following steps of: 1, acquiring a multi-modal aspect-level sentiment data set; 2, inputting a pre-training language model to obtain global context features; 3, combining target emotion semantic fusion to obtain aspect-related visual emotion features; 4, obtaining a dependency relationship among text information words, and obtaining syntactic enhanced text features; 5, fusing the global context feature, the syntactic enhanced text feature and the aspect-related visual emotion feature to obtain a multi-modal fusion feature; the method has the following beneficial effects that the feature information is expanded, so that effective interaction of text images is realized, noise generated by text and picture data is reduced, and the quality of multi-modal fusion features is ensured.

The invention discloses a Tibetan rumor model training method, a Tibetan rumor model detection method and a Tibetan rumor model detection device, and the Tibetan rumor model detection method achieves the remarkable beneficial effects through the following technology: firstly, encoding a Tibetan text by using a CINO pre-training model, generating an initial representation vector capable of capturing specific grammar and semantic features of Tibetan, and then performing classification on the initial representation vector; the problem that modeling is difficult when the Tibetan language serves as a low-resource language under the condition of data scarcity is effectively solved. Secondly, local semantic features are extracted through a convolutional neural network, and n-gram modes of different granularities in a text are captured by using a convolution kernel, so that the recognition capability of phrases and local modes is enhanced. Meanwhile, a multi-head self-attention mechanism further processes local features, models a long-distance dependency relationship and generates global context information, and the modeling difficulties that the Tibetan syntax is complex and the dependency relationship is far are solved.

The invention provides a complex question and answer method and system based on adaptive task deconstruction and multi-modal evidence aggregation, belongs to the technical field of natural language processing, and designs a dynamic Few-shot prompt construction method based on dependency syntax fingerprints to ensure that a prompt template is matched with a question structure; the invention discloses a dynamic problem deconstruction method based on confidence evaluation and auto-reflection. The method comprises the following steps: recursively decomposing a problem tree by using a large language model; converting the problem tree into a standardized linear task execution sequence by a problem tree context dependence specification and task sequence generation method; obtaining a high-correlation evidence set of each task based on an evidence generation method of two-way recall and cross encoder rearrangement; and the task sequence is reasoned and dynamically optimized by a question answer extraction method based on double-strategy aggregation reasoning. According to the method, the accurate complex question and answer result can be provided on the premise of ensuring the question disassembling quality, restraining error propagation and comprehensively recalling evidences.

The invention discloses an automatic service orchestration method, system and device and a storage medium, belongs to the technical field of computer science and information, and solves the technical problem that when an existing system processes complex queries and decisions, manual intervention is often needed, and the automation level is limited. The method comprises the steps that an intelligent agent obtains natural language query information input by a user and sends the natural language query information to the intelligent agent; preprocessing the natural language query information through the intelligent agent, wherein the preprocessing comprises word segmentation, part-of-speech tagging and syntactic analysis; the intent of the user is understood through the intelligent agent according to the preprocessing result, and a corresponding service process is planned in combination with a business rule and historical data; evaluating different service process options through the intelligent agent, making a decision, selecting an optimal scheme, and determining a specific execution path of the service process; starting to execute the service process through the intelligent agent according to the specific execution path; and confirming that all the steps of service orchestration are successfully executed through the intelligent agent.

The invention provides an intelligent number asking method for large model index query, and belongs to the field of natural language processing and database query. An auxiliary information base is constructed to store a database structure and historical question and answer examples, and key entities and fields are extracted by conducting structured format processing on natural language questions; a large language model guided by a small number of examples is utilized to generate an intermediate expression close to an SQL grammar and a plurality of SQL query candidates, and a consistency alignment mechanism is designed to verify candidate SQL in the aspects of field matching, aggregation functions, data tables and the like so as to screen out an optimal SQL query. Finally, the generated correct SQL and the corresponding question and answer pairs are fed back to the auxiliary information base, and self-learning enhancement is achieved. According to the method, under the condition that only a small number of training examples are needed, the accuracy and expandability of large language model text generation SQL query are improved.

Some embodiments perform fully unsupervised string error detection and repair, for strings containing both syntactic errors and semantic errors. In some embodiments, string error mitigation includes learning a regular-expression-based pattern that characterizes at least a predefined fraction of a set of string data values, identifying and masking a semantic substring, deriving edits to repair a string which contains the semantic substring, and replacing a mask value with a replacement substring that conforms to the pattern. Some embodiments utilize a language model. Some embodiments include or utilize a pattern learner, one or more learned significant patterns, and a repair engine. Some embodiments exploit program execution information to generate patterns when significant patterns are not initially found.

Computer-implemented methods and systems for processing geometry replacements are disclosed. The methods include decoding / encoding a syntax element associated with a coding mode from / into a bitstream associated with geometry displacements; and reconstructing / converting, based on a coefficient configuration associated with the coding mode, a plurality of quantized transform coefficients from / to a plurality of zero-run length codes.

The invention relates to a man-machine interaction dialogue method, system and device based on a natural language and a medium. The method comprises the steps that firstly, multi-modal interaction data is acquired and preprocessed, and segmented words and syntax are analyzed through a natural language processing technology to construct intention feature vectors; combining the intention feature vector with a historical dialogue record, and using a pre-trained language model to generate context semantic elements containing a semantic relationship; if the context semantic elements are matched with the preset scene feature library information, predicting a user intention change trend by adopting a reinforcement learning model to obtain an intention prediction result; and finally, evaluating user intention change based on an intention prediction result, extracting associated domain knowledge by utilizing a knowledge graph if significant change occurs, and inputting the associated domain knowledge into a dialogue generation model to obtain a natural language reply sequence. According to the method, the understanding precision of the user intention is improved, the dynamic prediction of the intention change is realized, the relativity and coherence of reply are guaranteed, and a more efficient processing path is provided for natural language man-machine interaction.

The invention provides a field NL2SQL data automatic synthesis method and system based on a large language model, and the method comprises the steps: building an SQL template library through collecting a public data set and real business query data, carrying out the grading according to the SQL complexity, and forming an SQL template set covering different difficulties; specific SQL queries are generated based on an SQL template library and a large language model, verification is carried out through multiple mechanisms such as grammar check, execution verification and result dimension consistency check, and correctness and performability of the SQL queries are ensured. Generating a plurality of candidate natural language questions according to the verified SQL query and the large language model, calculating a semantic consistency score of each candidate question and the SQL query through a cross consistency evaluation mechanism, screening out the question with the most consistent semantics as a final question-answer pair sample, and meanwhile, eliminating low-quality samples by setting a consistency threshold value, so as to obtain a final question-answer pair sample; and the data accuracy and consistency are further improved.

The invention discloses an intelligent SQL (Structured Query Language) generation method and system based on multistage intention recognition. The method comprises the following steps: receiving a natural language query of a user; vector matching: carrying out vector matching based on a knowledge base to obtain Top-K candidate query objects, the knowledge base containing multilayer structure description information; a large language model intention understanding step: determining a final query object in combination with the candidate query object description; loading a branch workflow, analyzing a field and a table structure, and dynamically configuring a database table and field information; sQL generation: generating SQL statements conforming to grammatical rules, including automatic field addition, time condition conversion and field priority matching operation; and executing the SQL statement and returning a result, and if the result fails, recording a log to prompt correction. By combining a natural language processing technology with a database knowledge base, an efficient, accurate and user-friendly database query solution is provided.

The invention relates to the technical field of text processing, in particular to a text compression method, a text decompression method, a model training method, devices and equipment.The text compression method comprises the steps that a to-be-compressed target text is obtained, and the to-be-compressed target text is input into a feature extraction module of a text compression model; using a deep learning technology to capture semantic and syntactic features of the text, and expressing the extracted key information in the form of an embedded vector sequence; and a q-former quantization processing module is used for carrying out quantization processing on the embedded vector sequence, and the continuous embedded vector sequence is converted into a discrete token sequence, so that effective compression of text information is realized. By adopting a form of combining a feature extraction module and a q-former structure, the redundancy of text data can be remarkably reduced, and meanwhile, the high-quality restoration capability of an original text is kept.

The invention provides a large model tool calling method and device based on multiple constraint sets and electronic equipment. The method comprises the steps that a current constraint set comprises a permission constraint set, a state constraint set and a context constraint set; based on the current constraint set and a tool set cue word builder, obtaining a tool set cue word corresponding to the current constraint set, and generating a calling intention based on the task demand instruction and the tool set cue word; and generating a tool call description based on a cue word formed by the large language model and the call intention, and executing tool function call under the condition that grammar and semantic check of the tool call description is passed. On one hand, a current constraint set uniformly manages authority, state and context constraint conditions, and the constraints are flexibly subjected to set operation to dynamically judge the availability of a tool; and on the other hand, the tool set cue words are dynamically screened, redundancy of the tool set cue words is avoided, the calculation amount of a large language model is reduced, and the possibility of calling wrong tools is reduced.