29 results about "Polysemy" patented technology

The invention discloses an automatic label labeling and classifying method and system oriented to unstructured system documents, and relates to the technical field of artificial intelligence. The method comprises the steps that semantic structure pre-analysis is conducted on an original system text, and a system semantic structure tree is constructed; establishing a system semantic enhancement vector space based on the semantic units and the logic relationship thereof; performing semantic deconstruction on the preset tag and extracting a feature vector; realizing cross-space semantic matching of the document and the tag through a system semantic attention mechanism; a confidence evaluation module is introduced to screen high-confidence labels from the three dimensions of structural integrity, coverage and logic consistency; and outputting a final label and a score through semantic conflict detection and resolution. According to the method, the problems that in the prior art, unstructured system text labeling accuracy is low and large-scale labeling samples are dependent on polysemy ambiguity, high context dependency, complex semantic structure and the like are solved, and labeling accuracy and robustness are remarkably improved.

The invention belongs to the technical field of ancient book digital processing, and discloses a training knowledge-fused ancient book word sense disambiguation deep learning method, which comprises the following steps of: constructing a training knowledge graph; preprocessing ancient book image texts; constructing a deep semantic disambiguation model fusing training knowledge; word sense disambiguation reasoning and result output; and system integration and intelligent application interface design. The high-precision word sense disambiguation method oriented to ancient books and texts is constructed by fusing training knowledge and a deep learning technology, so that the recognition capability of complex semantic phenomena such as polysemy words, ancient and modern heterosemy and common and false characters is remarkably improved, the interpretability and field adaptability of the model are enhanced, the dependence on manual training is reduced, and the training efficiency is improved. The method realizes efficient understanding and intelligent processing of the semantics of the ancient books under the condition of low resources, and has good popularization and application prospects and culture inheritance value.

The invention relates to the technical field of translation, and discloses a translation calibration method based on big data, which can accurately identify semantic deviation in a multi-meaning scene, for example, accurately judge that'cell 'should be translated into'cell' instead of'house 'in a biological text. The system can detect the language field mismatching problem of spoken expression in the official file, and provides term correction suggestions conforming to industry standards. Aiming at technical terms with multiple meanings in legal clauses, the calibration process can be combined with context window analysis to select an optimal translation scheme, the omission ratio of manual review is remarkably reduced, continuous optimization and cross-language knowledge sharing of a translation calibration model are realized, and the accuracy of translation calibration is improved. The problems that a traditional system is lagged in updating and low-resource language performance is insufficient are solved.

The invention discloses a typhoid theory teaching method based on a knowledge graph, and belongs to the technical field of typhoid theory teaching. The method comprises the steps of multi-version data acquisition and standardization, ancient Chinese context semantic unit construction, term semantic disambiguation based on comparative learning, semantic conflict detection based on consistency constraint, multi-source information fusion and map optimization and teaching path generation. Through ancient Chinese context semantic modeling and comparative learning, precise disambiguation of traditional Chinese medicine ancient book terms is realized, and knowledge errors caused by polysemy of one word are reduced; through an automatic detection mechanism of logic consistency constraint, the quality and reliability of the knowledge graph are improved; a version traceability and belief propagation optimization strategy is adopted, so that the finally constructed knowledge graph not only can integrate multi-version essence, but also can clearly trace knowledge sources and evaluate knowledge credibility; the teaching path chain generated based on the atlas can dynamically simulate the whole process of traditional Chinese medicine syndrome differentiation treatment, and the clinical thinking training effect is remarkably improved.

The invention discloses an ancient modern text machine translation method based on deep learning, and belongs to the technical field of natural language processing. The invention provides a multi-task collaborative optimization framework aiming at the problems of domain knowledge deficiency, frequent word activation, corpus scarcity and the like of an existing neural network model in ancient language translation. Precise domain knowledge injection is realized by constructing a closed ancient language parallel corpus retrieval library and combining historical term alignment and logic chain reasoning; a probability weighted part-of-speech embedding mechanism is adopted, and grammar feature distribution of virtual word ambiguity and word type activity is dynamically fused, so that the grammar analysis robustness is improved; and semantic generation, retrieval alignment and part-of-speech constraint tasks are jointly optimized through a fixed weight strategy. According to the method, the semantic accuracy and the logic continuity of ancient language translation are improved, and extended application of complex ancient language styles is supported. Experimental results verify that the performance of the method is improved in terms of BLEU and ROUGE-L indexes, and efficient technical support is provided for ancient book digitization and cross-time explanation.

This invention relates to the field of web log analysis, and more particularly to a method and system for web log analysis based on big data. The method includes: calculating the weight of each decision tree in a pre-defined optimal random forest model; inputting real-time acquired web logs into the optimal random forest model; and using the weights of each decision tree to perform a weighted summation of the prediction results to obtain the final prediction result for the web logs. This invention fully considers the phenomena of polysemy and polyphy when analyzing web logs using keywords, facilitating the further integration of log prediction results based on keyword importance to obtain the final prediction result, thus improving the accuracy of the prediction results.

The application relates to the field of data processing, in particular to an AI semantic analysis and data processing method based on a large language model. First, professional terms and context windows in vertical field texts are extracted, hidden layer vectors are extracted by using the large language model, and spatial dispersion is calculated, and the terms are divided into three-state working modes according to semantic stability. For polysemous terms, independent semantic branches are identified through vector clustering, co-occurrence word matching and vector distance determination are used to realize accurate semantic attribution in a new context, and dynamic segmented adjustment of static term frequency (TF-IDF) weight is carried out accordingly, and finally the improved weight is injected into the downstream analysis process. The application gives the same term different weight expressions in different contexts, and the low-overhead cascade disambiguation mechanism efficiently solves the polysemy problem of terms, avoids confusion of structured data and error clustering of documents, and significantly improves the accuracy of text matching, information extraction and document classification and the like.

The application discloses a kind of information retrieval method and system based on large language model topological semantic reordering, including Euclidean rough screening, topological gravity calculation, power law sorting and result output step.This method uses vector embedding model to carry out rough screening to query and candidate document, then calls large language model to obtain the logarithmic probability value of each candidate document, and is re-sequenced according to power law decay law.The application also discloses a recall boundary determination method and a model alignment degree detection method based on semantic decay index.In polysemy disambiguation task, retrieval accuracy can be improved from 69% to 99%, with the advantages of high retrieval accuracy and self-adaptive model alignment state.

The invention relates to a risk content identification method and device, a storage medium and electronic equipment, and relates to the technical field of machine learning, and the method comprises the steps: obtaining to-be-identified resource content and to-be-detected resource comment information corresponding to the to-be-identified resource content; based on a preset indirect semantic keyword, matching target resource comment information containing the preset indirect semantic key from the to-be-detected resource comment information; performing semantic recognition on the target resource comment information based on a preset polysemy semantic classification model to obtain a comment detection result; and determining whether the to-be-identified resource content belongs to risk resource content according to the comment detection result. The accuracy of the identification result of the risk content is improved.

The present application relates to the technical field of data governance and artificial intelligence, in particular to a project attribute data intelligent auditing method and system, the method comprising: constructing a domain knowledge base containing dictionaries of project regions, types and the like, and a polysemy and synonym mapping table; obtaining a project name text to be audited, using a hybrid semantic analysis model based on the knowledge base to extract dimension information such as project regions, years and confidence levels; obtaining corresponding structured attribute values; arbitrating attribute missing, errors or to-be-confirmed through a rule engine combined with confidence levels to determine, and outputting results containing problem dimensions, suggested values and the like; generating work order update data, and feeding back effective mapping optimization of the knowledge base. The system comprises knowledge base management, semantic analysis, data interface, rule engine, closed loop processing and visual display modules, and each module cooperates to realize intelligent auditing. The present application can improve the intelligence and automation level of auditing, realize self-learning, reduce labor costs, and effectively guarantee project data quality.

The application discloses a decoding method and device of a Huffman coded image and a display device. Whether the current group of data to be decoded in a data buffer is a polysemy is judged. If not, the current group of data to be decoded is compared with one of a plurality of minimum code word extension values in sequence. If the current group of data to be decoded is smaller than a certain minimum code word extension value, the comparison is stopped, a previous code word extension value of the certain code word extension value is taken as a first minimum code word extension value, a first real code length corresponding to the first minimum code word extension value is found in a mapping relationship, and bit data with the first real code length is read from the current group of data to be decoded as a first Huffman code word. In the embodiment of the application, the data to be decoded is first distinguished, the real code length of the current group of data to be decoded is determined after it is determined that the data to be decoded is not a polysemy, the number of invalid comparisons in the decoding process is reduced, and the decoding efficiency is improved.

The application provides a Chinese named entity recognition method based on an attention mechanism, an electronic device and a storage medium, which comprises the following steps: inputting a text to be recognized into an embedding layer to obtain a word vector; using a Transformer encoder to extract features of the word vector to obtain a first context feature; using a Bi-LSTM model to extract features of the word vector to obtain a second context feature; fusing the first context feature and the second context feature to obtain a fused feature; and decoding the fused feature to obtain Chinese named entities corresponding to the text to be recognized. The Chinese named entity recognition method based on the attention mechanism realizes deep fusion of global semantic information and directional information. In order to obtain more context information and solve the problem of polysemy, a RoBERTa-wwm pre-training model is used as a character-level embedding, so that the model recognition effect is improved.

The invention relates to the field of data processing, in particular to a traditional Chinese medicine vertical domain large model training method based on a knowledge graph. According to the method, the symptom keyword library associated with the syndrome keywords is systematically constructed, the easy-to-mix tuple library is formed based on clustering of the symptom keyword library, marking and context disambiguation are performed on the syndrome keywords based on the easy-to-mix tuples in the sample corpus traversal process, and the marked keywords are replaced according to the matching result, so that the accuracy and the reliability of the system are improved. And carrying out iterative verification and optimization on the easy-to-mix tuple by means of a replacement keyword library, and finally obtaining high-quality and high-consistency corpora as pre-training data. According to the method, automatic processing and semantic optimization of corpora in the field of traditional Chinese medicine are realized, the semantic ambiguity problems of one-word polysemy, multi-word synonymy and the like in the traditional Chinese medicine text are effectively solved, the semantic consistency of knowledge graph and language model training is improved, and a high-quality training basis is provided for a large model in the field of traditional Chinese medicine.

The invention relates to the technical field of data management and artificial intelligence, in particular to an intelligent auditing method and system for project attribute data, and the method comprises the steps: constructing a domain knowledge base containing dictionaries of project regions, types and the like and polysemy and synonym mapping tables; obtaining a to-be-audited project name text, and extracting project region, year and other dimension information and confidence by using a hybrid semantic analysis model based on the knowledge base; obtaining a corresponding structured attribute value; judging attribute missing, errors or to-be-confirmed through a rule engine in combination with confidence arbitration, and outputting a result containing problem dimensions, suggested values and the like; and generating work order update data, and feeding back the effective mapping optimization knowledge base. The system comprises a knowledge base management module, a semantic analysis module, a data interface module, a rule engine module, a closed-loop processing module and a visual display module, and all the modules cooperate to achieve intelligent auditing. The auditing intelligence and automation level can be improved, self-learning is realized, the labor cost is reduced, and the project data quality is effectively guaranteed.

The invention discloses an AIGC-based culture feature accurate identification method. The method comprises the steps of obtaining original data; performing cultural image feature extraction on the original image through a pre-trained AIGC semantic generation model to generate a corresponding initial semantic candidate text set; performing spatial consistency matching and time consistency matching with the initial semantic candidate text set by utilizing geographic information and spoken semantic text data in original data to obtain a matched first semantic candidate text set; performing semantic similarity matching on dictated semantic text data in the original data and the first semantic candidate text set to construct a final second semantic candidate text set; and constructing a corresponding semantic path diagram based on the second semantic candidate text set and the corresponding original data. The invention further provides a cultural feature accurate recognition system based on the AIGC. The method provided by the invention can effectively solve the ambiguity judgment problem in cultural graph recognition.

The invention relates to a live broadcast fragment editing and recommending method and system based on cross-modal understanding, and belongs to the field of video editing. The method comprises the following steps: collecting original live broadcast stream data to obtain a feature vector sequence; constructing an interest hypergraph based on the feature vector sequence to obtain an activation intensity vector; performing cross-modal gradient calibration based on the feature vector sequence to obtain an aligned new feature vector sequence; based on the new feature vector sequence, performing ruminant attention suppression to obtain a fusion feature; ambiguous highlight judgment is carried out based on the activation intensity vector and the fusion features, a series of highlight moments are obtained, and each highlight moment is attached with the dominant interest dimension and the score of each dimension; and performing multi-version intelligent editing and personalized recommendation based on the highlight moment. The method is an intelligent method for understanding ambiguity of live broadcast content, aligning modal time, inhibiting modal prejudice and generating multi-version clips.

The invention discloses a human motion prediction method and device based on an intention understanding graph convolutional network, and the method and device have the following beneficial effects: through constructing a dynamic intention adjacency matrix based on speed similarity, a semantic association relationship which changes with time between joint nodes can be captured; the limitation that a traditional static adjacent matrix cannot reflect the action intention is overcome, and the model can understand the potential motivation behind the action; an intention understanding module is introduced to be combined with a space-time diagram convolutional network, joint feature modeling of a semantic layer and a motion layer is achieved, the accuracy and semantic consistency of polysemy action prediction are effectively improved, and prediction ambiguity is reduced; intention feature aggregation is carried out by adopting a graph attention network, so that neighbor node weights can be adaptively allocated, the contribution of key intention features is highlighted, and noise interference is reduced; the spatial-temporal features and the intention features are integrated through a weighted fusion strategy, and a prediction sequence is generated by using a time convolutional network, so that the continuity, smoothness and stability of a prediction result are ensured.

The invention relates to a polysemy content display method and device for electronic learning equipment and electronic equipment, and the method comprises the steps: building and storing a mapping relation between at least two word meaning items of a target word and corresponding example sentences, and enabling each word meaning item to be uniquely associated with at least one corresponding example sentence; a selectable list containing at least two word meaning items is displayed in a first area of a display interface of the electronic paper display screen, and the list is provided with a currently activated visual focus indication; displaying example sentences corresponding to the word meaning items identified by the visual focus indication at present in a second area of the display interface of the electronic paper display screen; in response to a received selection instruction for a target word meaning item in the selectable list, switching the visual focus indication to the target word meaning item; in response to the switching indicated by the visual focus, the content displayed by the second area is synchronously and uniquely updated. According to the method, the cognitive burden of the user among different display contents is eliminated, and the concentration degree, the understanding accuracy and the operation efficiency of word learning are improved.

The invention discloses a key process identification method and device for a power text, equipment and a storage medium, and the method comprises the steps: obtaining the power text of a to-be-identified key process, carrying out the semantic representation of the power text, generating a semantic vector, analyzing the depth meaning of a sentence through the context semantics of each sentence of the power text, and obtaining the key process of the power text; and performing semantic enhancement on the semantic vector to obtain an enhanced semantic vector, and identifying the key process of the power text based on the enhanced semantic vector. Therefore, the context semantic analysis is used for mining the depth meaning of the sentence, the deep semantics of the power operation word, the sequential logic of the operation steps and the entity equipment association information can be accurately captured, the semantic vector representation accuracy is remarkably improved, the key process identification accuracy is greatly improved, the method is adaptive to complex scenes such as polysemy and synonyms in the power text, and the method is suitable for popularization and application. And the operation specification and efficiency are ensured.

The application relates to a semantic similarity-based question and answer system search matching method and application thereof. The question and answer system comprises standard questions and answers, and a plurality of similar questions are set corresponding to the standard questions; a sample set is constructed, a pre-training model is adjusted and trained; questions in the question and answer system are vectorized, new user questions are vectorized, cosine similarity calculation is carried out, all questions in the question and answer system are sorted according to the similarity scores, and a preset number of questions are output; the method is applied to a tax question and answer system. The application solves the problems of polysemy, word ambiguity, word segmentation accuracy and the like of a traditional model, can fully obtain semantic information contained in a sentence, greatly improves the performance of the model, the training and prediction stages are fully parallel, the prediction accuracy of the model is improved, search recommendation indexes are improved, the generalization ability of the model is enhanced, user experience is improved, and maintenance cost is reduced; and the application is particularly suitable for the field of tax consultation.

The application discloses a kind of based on BIM model's component and the quick search method and system of associated relationship, which method includes the following steps:1, initialize the core configuration parameter of search engine;2, initialize all matrix information;3, floor information is loaded into matrix;4, component information is loaded into matrix, and component index table is established;5, user imports polysemy line or component information, and start searching;6, according to polysemy line information, directly accurately calculate search matrix;7, obtain very few components from search matrix;8, carry out de-duplication fault-tolerant processing to component;9, component is carried out six-layer quick filtering and screening;10, return the result searched.This application relates to the technical field of building engineering, and can solve the technical problems existing in prior art.

The invention provides a text parsing agent system and method based on natural language processing, belongs to the technical field of computer data processing, is used for deep semantic understanding and automatic logical reasoning of complex texts and generating personalized executable suggestions, and comprises the following steps: receiving to-be-parsed text data; the to-be-analyzed text data is preprocessed to generate a semantic knowledge joint vector set, a temporary logic diagram is constructed based on the semantic knowledge joint vector set, and the temporary logic diagram represents the logic relation of the to-be-analyzed text data. According to the method, iterative semantic disambiguation and knowledge verification are combined, so that the defect of error accumulation in a traditional linear analysis process is overcome, semantic hypothesis and background knowledge can be subjected to two-way interaction and mutual calibration, the analysis precision of complex sentence patterns, terminologies and polysemy words is remarkably improved, and the analysis efficiency is improved. And deeper and more reliable text semantic understanding can be obtained.

The invention discloses a private data desensitization optimization method and system for document archives, and belongs to the technical field of natural language processing. According to the method, a semantic space heterogeneous graph fusing text semantic logic and a layout space structure is constructed, and an unstructured document file is converted into a computable topological structure; an anti-ambiguity sensitive entity identification method combining space correction attention and document topic constraint is adopted to realize accurate positioning of polysemy sensitive entities; a self-adaptive generalization instruction is generated based on information entropy risk assessment and access permission constraint, and a differential desensitization strategy is dynamically matched; and finally, through a desensitization data flow recombination mechanism, completing sensitive information replacement and format repair on the premise of keeping the original layout structure and semantic coherence. According to the scheme, the problems of large semantic ambiguity, desensitization strategy stiffness, format distortion after desensitization and the like in an existing document desensitization process can be effectively solved, and the accuracy, availability and compliance of document file desensitization processing are improved.

The application discloses a kind of based on weight cooperation, multi-modal semantic fusion and knowledge graph anchoring key word extraction method and system, belong to natural language processing technical field.The method first introduces the statistical weight of candidate term by the improved TF-IDF algorithm of context perception factor, and generates dynamic context word vector in combination with ELECTRA pre-training model to capture one word polysemy;Further construct term relationship graph network, utilize graph convolutional neural network modeling complex topological relationship between terms;Innovatively propose three-level fusion mechanism, and introduce knowledge graph anchoring module to map term to domain knowledge base for professional verification and dynamic expansion.The application effectively solves the problems of insufficient semantic understanding, poor field adaptability and uninterpretable decision of traditional methods, and significantly improves the accuracy and recall rate in the key word extraction task of medical, legal and other professional texts.

The present application relates to the field of intelligent dialogue, and relates to a multi-task learning dialogue method supporting anaphora resolution and polysemy, wherein an anaphora resolution model, a polysemy model and a multi-task learning model are trained through training corpus, historical dialogue and current dialogue are spliced into context representation text, anaphors, zero pronouns and anaphora in the text are labeled through the anaphora resolution model, the anaphora probability of the anaphors and the anaphora probability of the zero pronouns and the anaphora are calculated, and whether the anaphora can be resolved is judged according to a set threshold, if the anaphora can be resolved, the anaphors or the zero pronouns are replaced or supplemented into corresponding anaphora to form a replacement text, the intent key components in the replacement text are extracted through the polysemy model, the domain, intent and slot of each intent key component are obtained through the multi-task learning model, and are sent into a dialogue management module to obtain the dialogue action at this moment, so that the intent can be accurately understood.

The application discloses a kind of private data desensitization optimization methods and systems for archives, belong to natural language processing technical field.The method is by constructing the semantic space hetero-graph that fuses text semantic logic and layout space structure, it is converted into the topological structure that can be calculated to unstructured archives;Adopt the disambiguation sensitive entity recognition method that combines space correction attention and document theme constraint, realize the accurate positioning of polysemy sensitive entity;Based on information entropy risk assessment and access permission constraint, generate adaptive generalization instruction, dynamically match different desensitization strategies;Finally, through desensitization data stream reorganization mechanism, complete sensitive information replacement and format repair under the premise of maintaining original layout structure and semantic coherence.The scheme can effectively solve the problems of semantic ambiguity, desensitization strategy rigidity and layout distortion after desensitization in the existing document desensitization process, improve the accuracy, availability and compliance of document archive desensitization processing.

The application discloses an event detection method and device based on semantic network word representation and attention graph, comprising the following steps: splicing the word content vector, the word structure vector and the position feature vector of each word to generate the feature map of each sentence; combining the POS vector of each word to calculate the attention mechanism and generate the new feature map of each sentence; generating the sentence-level feature vector based on the new feature map; and obtaining the event detection result by using the splicing result of the sentence-level feature vector and the word content vector. The application comprehensively utilizes the external corpus, the semantic network, the part of speech and the attention graph, optimizes the features, more accurately extracts the trigger word, introduces more information, solves the polysemy problem, expresses the associated information between the synonymous words, and obtains more accurate event detection result.

The application relates to the field of artificial intelligence translation technology, and provides an artificial intelligence translation optimization device for professional terms, which comprises a translation device main body, a touch display screen for displaying an interactive interface and a key for inputting operation instructions are sequentially arranged from back to front at the top end of the translation device main body, and a sound collecting hole is fixedly connected to the front and back sides of the top end of the translation device main body; the sound collecting hole on the back side is arranged at the back end of the display screen to collect the voice of the opposite party, and the sound collecting hole on the front side is arranged at the front end of the key to collect the voice of the party, and a sound amplifying hole is arranged on the left end back side and the right end front side of the translation device main body. Through visual scene and speech text cross-modal fusion disambiguation, combined with scene semantic constraint vector and domain knowledge graph vector, weighted matching calculation is carried out, so that the polysemy ambiguity problem of professional terms can be accurately eliminated, and the scene adaptability and translation accuracy of professional term translation are significantly improved.

The invention provides an industrial product parameter extraction system, which comprises a knowledge modeling and learning module used for receiving field expert input through a human-computer interaction interface and constructing a structured field knowledge model based on a natural language processing technology; the model training module is used for generating training data based on the domain knowledge model and performing supervised training or fine tuning on the deep learning extraction model; the parameter extraction application module is used for receiving a to-be-processed document, performing parameter extraction by adopting the trained deep learning extraction model, and performing standardization processing on an extraction result; and the knowledge base module is used for storing the domain knowledge model and the extracted structured parameter data. According to the method, the extraction accuracy and the semantic comprehension capability are remarkably improved, and particularly, the method is excellent in performance when being used for processing complex tables, polysemy words and heterogeneous expressions.