39 results about "Lexical semantics" patented technology

The invention relates to an entity relationship extracting method of the Zang language. The method comprises the following steps: extracting training linguistic data from the Zang-Chinese text linguistic data information; constructing a Zang word vector model; acquiring an entity relationship characteristic vector from the Zang word vector model; using the entity relationship characteristic vector as an input to construct an entity relationship classification model based on a neural network; and applying multiple layers of characteristic extractions to the entity relationship characteristic vector, thereby finally acquiring a Zang language entity relationship classification. The extraction of the Zang language entity relationship is achieved by constructing the Zang word vector model, researching and solving lexical semantic characteristics and sentence characteristic vector expression methods of the Zang language entity relationship, and further constructing the Zang language entity relationship classification model, accordingly increasing the accuracy in the Zang language entity relationship classification, and providing technical supports and services to the researches in the fields of the Zang language knowledge mapping, question-answering system, information extraction, information search, and the like.

The invention discloses an extraction method of a semantic relation between Chinese entities. The extraction method comprises the following steps of: carrying out syntactic analysis on natural statements to determine a complete syntactic tree of the natural statements; extracting a shortest path containing tree between two Chinese entities from the complete syntactic tree; extracting a path verb nearest to a second Chinese entity from the shortest path containing tree; respectively acquiring the semantic information of the two Chinese entities and the path verb; adding the three acquired semantic information into a root node of the shortest path containing tree according to a preset rule to determine the expanded shortest path containing tree to be a natural statement relation tree; and carrying out relation classification on the relation tree by utilizing a prestored classification model. According to the extraction method of the semantic relation between Chinese entities, which is disclosed by the invention, the relation tree contains abundant structured information and lexical semantic information and has better generality and semantic relation extraction overall performance, the dependence degree of a large-scale corpus is relieved, and meanwhile, the calculated amount of the system is lower.

A lexical semantic structure for modeling semantics of a natural language input on a computer is described. A set of lexical semantic categories is selected to model content of the natural language input. A methodology associates content of the natural language input to one or more categories of the set of lexical semantic categories.

The invention discloses a method and system for mining Chinese event information. The method includes the steps of defining an event sample model based on the mining requirement, instantiating the event sample model to obtain a seed event, taking the seed event as the foundation of an event mined from an original text, conducting preprocessing on the original text to obtain a candidate template set, conducting information labeling on the seed event, processing the seed event on the basis of labeling information to obtain a seed template set, then, processing the candidate template set through an iterative method according to the grade scores of candidate templates and the semantic similarity between the candidate templates and seed templates so that an event mining anchor set can be obtained, and obtaining the event types corresponding to mining event anchors according to the lexical semantic similarity between each event mining anchor in the event mining anchor set and any seed event anchor in the seed event set. The aim of mining the Chinese event information from the original text can be achieved only by labeling a small amount of samples, namely, the seed events, and the mining cost is reduced.

A word semantic embedding apparatus includes: a data storing unit to store a lexical semantic network including word dictionary data and word semantic data; a word list generating unit to extract vocabularies having a predetermined part of speech from the stored word dictionary data and generate a list of words to be learned; a processing data generating unit to bring the generated list of words to be learned and the word semantic data of a word to be learned included in the list of words to be learned from the data storing unit and process the data suitable for word embedding learning to generate processing data; and a word embedding learning unit to learn the word to be learned through the word embedding learning using a learning model formed of an input/output layer and a projection layer with the generated processing data to generate a word vector.

The invention discloses a Chinese-English cross-language vocabulary representation learning method and system based on paraphrasing primitive words, which represents the vocabulary of Chinese and English languages in vector form in the same vector space, and obtains more accurate word embedding combined with semantic information. Firstly, the set of paraphrasing primitives is obtained by processing the paraphrasing relations in Chinese dictionaries, so that the words in the set of paraphrasing primitives can cover all the lexical semantics in dictionaries. Secondly, all the words in Chinese dictionaries and English dictionaries are represented by the vectorized expressions of the paraphrasing primitives. The vectorized expressions of the paraphrasing primitives are used to express all thewords in the dictionaries. Finally, combining with the context and semantic relationship of Chinese and English corpus, we set certain weights on the expression of paraphrase primitives in vocabularyto obtain more accurate semantic relational word embedding. Compared with the existing word embedding, the invention has the advantages of high word embedding accuracy, strong expansion ability, convenient realization and the like, and can better serve the subsequent natural language processing tasks.

The invention provides a psycholinguistic evaluation method and system for Chinese aphasia. The method comprises performing auditory analysis testing, speech input buffer testing, speech input dictionary testing and lexical semantic testing on a patient in sequence; performing visual discrimination testing, concept semantic testing, lexical semantic testing, speech output dictionary testing and speech motion planning testing on the patient in sequence; sequentially performing glyph visual recognition testing, glyph input dictionary testing, and lexical semantic testing on the patient in sequence; and evaluating and judging evaluation test results. The system includes a main server, a client and a human-computer interaction terminal. The main server includes a database and a listening comprehension evaluation process subsystem, a graph naming evaluation process subsystem, a reading evaluation process subsystem and a test evaluation module. According to the psycholinguistic evaluation method and system for Chinese aphasia, the damage of different properties in aphasia can be determine by evaluation testing, and a therapist can be assisted to accurately locate the damaged part of thebrain region, thereby achieving targeted treatment of aphasia and improving the treatment efficiency.

The invention provides a text feature extraction method based on semantic analysis. The method comprises the following steps: initializing a text corpus, performing word segmentation and unused word removal processing on texts, clustering the texts according to semantic relevance (the formula is as shown in the specification), constructing a lexical semantic network model graph, obtaining the status and the contribution of a vocabulary in the whole text according to the importance of the vocabulary in a lexical semantic network model, setting a proper importance threshold, and extracting a feature vocabulary vector of the text. Compared with the tradition word frequency-inverse document frequency method, the text feature extraction method provided by the invention has the advantages of having higher accuracy, overcoming the defect that only one category of text features can be extracted by the information gain method, having higher application value, accurately calculating the contribution of different vocabularies to the idea of the text and meanwhile providing good theoretical basis for subsequent text clustering.

The invention provides a big data processing method based on machine learning, which comprises the following steps: given a retrieval sentence, filtering the words in the initial retrieval by using auniversal stop word list, and reserving meaningful retrieval words; the semantic block model is used to represent the lexical semantic vector. On the basis of semantic vector, cosine similarity is used to find out the closest words from other words for each initial search term, which can be used as extended search terms. The corresponding extended search terms in the initial search are used to replace the original search terms, and the newly generated search term sequence is used as the extended search terms. According to the permutation and combination of the extended search terms, the extended search sentences with different expressions are obtained. The invention improves the parallel frame of the MAPRUDUCE and is better adapted to the needs of text data mining. Aiming at the irregularity of social text, semantic vectors are used to represent and analyze the text data effectively, which is suitable for the analysis and calculation of social text mining of various scales.

The invention provides an ontology concept-based lexical semantic similarity solving method, which comprises the steps of mapping to-be-compared words input in a statistical method module into an ontology concept; selecting ontology concepts, with corresponding maximum depths, of the to-be-compared words from an ontology concept module, calculating the distance between the ontology concepts and calculating the most recent common ancestor depth; and finally calculating the similarity between the two to-be-compared words. The ontology concept-based lexical semantic similarity solving method is closer to an empirical value of an expert in quantitative concept; the factors of the distance between the ontology concepts, with the corresponding maximum depths, of the to-be-compared words (c1, c2), the depths and the like are more fully and comprehensively considered, so that the accuracy of the semantic similarity result is greatly improved; and the ontology reasoning effect is better improved.

The invention discloses a method for realizing text feature extraction based on an improved small-world network model. According to the method, a semantic relevancy function is determined according to a Chinese word segmentation preprocessing process and determined vocabulary position weights and word class weights in combination with a (HowNet) two-vocabulary relevancy algorithm and a vocabulary-to-text importance method, wherein the function is subjected to normalization processing, and calculation conditions of values are more standard; and two parameters, namely a density parameter and a weight parameter are set for a lexical semantic network model graph, the two parameters are effectively fused, and an appropriate threshold value is set to extract text feature vocabularies. The method has higher accuracy and overcomes the defect that a traditional method is only suitable for extracting text features of one category; the method has higher application value, contribution degrees of different vocabularies to text thought can be precisely calculated, data processing is more standard, the result error rate is lowered, the constructed lexical semantic network model graph better conforms to the actual condition, and meanwhile a good theoretical basis is provided for subsequent text clustering.

The invention discloses a mining and displaying method combining vocabulary matching extraction and semantic classification, and belongs to the field of natural language processing and language learning application. The method comprises: extracting and filtering word collocation by using a deep learning-based dependency syntax semantic model and a statistical method; in combination with the sorted word classification and the extracted collocation, calculating a many-to-many collocation matrix; screening, sorting and recombining the collocation words and the word classification according to the collocation density; displaying the finally obtained word classification matching matrix in an m*n or m*n*o card form; and the cards are displayed according to a hierarchical sequence of lexical semantic classification. Through the method, a collocation and vocabulary classification system with more general significance and a many-to-many word classification collocation matrix display form can be obtained, the display form is visual and vivid, the information density is large, and systematic learning of language vocabularies is facilitated through a semantic and pragmatic combination grouping mode.

The invention provides a word vector correction method based on semantic relation constraint and a computing system. The method comprises the following steps: a synonym and antonym constraint set with a symmetric relation and a direct hypernym/hypernym constraint set with an asymmetric relation are extracted as an external knowledge source update word embedding vector from WordNet and Roget semantic dictionaries; and the similarity of the pair of vocabularies can be calculated by randomly inputting two vocabularies of which the similarity needs to be calculated. The vocabulary semantic similarity calculation system comprises an input unit, an initialization unit, a calculation unit and an output unit. According to the method, the existing neural network word embedding vector is updated based on the lexical semantic relationship constraint provided by the external knowledge source so as to be used for lexical semantic similarity calculation. The updating speed and the word embedding vector semantic updating effect are obviously better than those in the prior art, and the accuracy rate in calculating the word semantic similarity is higher.

The invention provides an improved lexical semantic similarity solution algorithm. On the basis of a successfully initialized statistical method module, a maximum keyword is extracted by comparing context words with the maximum weights of words to be compared and the similarity of the words to be compared, and finally the similarity between the extracted maximum keyword and the words to be compared is calculated to obtain a result. The calculation result of the semantic similarity is basically consistent with actually manually judged semantic similarity; the objective reality is better reflected; and the user demands are better satisfied.

The invention provides a novel method for solving the lexical semantic similarity between ontology concepts. The method comprises the steps of calculating the similarity between the ontology concepts, with the maximum depths, of to-be-compared words input into a statistical method module; calculating the word form similarity between the two to-be-compared words, calculating the influence of the most recent common ancestor depth of the two to-be-compared words on the similarity between the two to-be-compared words and constructing an impact factor function; and finally calculating the similarity between the two to-be-compared words. The novel method is closer to an empirical value of an expert in quantitative concept; the factors of the distance between the ontology concepts, with corresponding maximum depths, of the to-be-compared words, the depths, the density and the like are more fully and comprehensively considered, so that the accuracy of the semantic similarity result is greatly improved; the ontology reasoning effect is better improved; better improvement of the accuracy of the word form similarity result and the semantic similarity result between the words is considered; data processing of various influence factors is more standard; and the novel method accords with the practical application effect.