359 results about "Word meaning" patented technology

It is an object to highly accurately perform semantic relationship extraction from text data by performing supervised learning of multiple classes using an existing thesaurus as a correct answer. Concerning any pair of words in a text, a plurality of kinds of similarities are calculated and a feature vector including the similarities as elements is generated. A label indicating a classification of a semantic relationship is given to pairs of words on the basis of the thesaurus. Data for semantic relationship identification is learned as an identification problem of multiple classes from the feature vector and the label. Identification of an inter-semantic relationship of two words is performed according to the data for semantic relationship identification.

A system, method, and computer program product for natural language understanding, including forming pairs of words and/or phrases in independent phrases in a sentence, wherein each word and/or phrase is paired with every other word and/or phrase; determining meanings for the words and/or phrases; comparing the meanings of each word and/or phrase with each meaning of the other word and/or phrase in the pair; selecting the pairs with the best relationships; combining highly-related pairs with other highly-related pairs to form longer groups of words; exchanging word meanings for the longer groups with word meanings having a weaker relationship to determine if the exchanged word meanings provide an overall stronger relationship; and forming longer and longer groups until the sentence is understood.

The present invention provides a new method for indexing a given text objects, using text parsing module and words indexing databases. According to this method each word is assigned a first index code according to words meaning, a second index code according to each word syntax category and a third index code according to word syntactical role. The words indices are arranged according to hierarchical order based on syntactical relations between the text words. At the last stage, differentiating symbols, which represent indices hierarchical order, are assigned between adjacent words indices. The indexing process may be implemented as automatic computerized program or as wizard application enabling human intervention in the indexing process. The indexing method can be utilized for enabling text search utilities based on matching between The query indices and source text indices.

The invention relates to the technical field of man-machine interaction, and discloses an information processing method and device for realizing intelligent question answering. The information processing method comprises the following steps of: carrying out sentence segmentation on question text information to obtain a user question; and searching a standard question most similar to the user question and corresponding answer information from a QA library on the basis of a question similarity. Compared with the existing keyword retrieval-based question answering method, the method disclosed by the invention does not need to require the users to have keyword decomposition ability, is automatic in the whole process and is capable of greatly enhancing the user experience and improving the search effect and the pertinence and effectiveness of answers. Meanwhile, through fusing natural language understanding technologies such as sentence model analysis, lexical analysis and lexical meaning extension, and carrying out comprehensive calculation on multi-dimensional similarity, the method is capable of improving the correctness of a final sentence similarity in a Chinese automatic question answering process, and enabling a Chinese intelligent question answering system to be possible.

A method and a system for automatically finding one or more answers to a natural language question in a computer stored natural language text database is disclosed. The natural language text database has been analyzed with respect to syntactic functions of constituents, lexical meaning of word tokens and clause boundaries, and the natural language question comprises a question clause. A computer readable representation of the question clause is analyzed with respect to syntactic functions of its constituents and the lexical meaning of its word tokens. In response to the analysis a set of conditions for a clause in the natural language text database to constitute an answer to the question clause is defined. The conditions relate to the syntactic functions of constituents and the lexical meaning of word tokens in the clause. Furthermore, clauses that satisfy said conditions are identified in the natural language text database, and answers to the question clause is returned by means of the identified clauses that matches the conditions.

The invention provides a multi-subject extracting method based on semantic categories. The multi-subject extracting method based on the semantic categories comprises the following steps that firstly, a document is preprocessed according to a traditional method and a vector composed of feature words is obtained preliminarily; secondly, synonyms are merged by the utilization of the corresponding relation between word meanings and concepts of 'HowNet', polysemic word disambiguation is carried out according to the correlation between the semantic categories and the context, and a concept vector model is constructed to represent the document; then the concept vector model is converted to be a semantic category model according to the one-to-one corresponding relation between the concepts and the semantic categories; the concept similarity is calculated by the utilization of the related semantic information in the concepts in 'HowNet' and then the semantic similarity is obtained; the semantic categories are clustered by improving the K-means algorithm according to the method of presetting seeds, and a plurality of subject semantic category clusters are formed; finally, a plurality of sub-subject word sets are obtained in a reverse mode according to the corresponding relations between the semantic categories and the concepts and between the concepts and words. The method considers the semantic information, overcomes the defect that the sensibility to the initial center by the K-means algorithm and time-and-space cost are not stable, and improves the quality of extracted subjects.

The invention discloses a CRFs (conditional random fields) and SVM (support vector machine) based method for extracting fine-granularity sentiment elements in product reviews. The method comprises the steps as follows: a, a CRFs model is adopted, review language characteristics are taken as sequences, then position labelling is performed on review languages according to the sequences, corresponding rules are adopted to perform stratified filtering on error labels, and extraction for sentiment subjects and sentiment words is finished; and b, an SVM model is adopted to perform sentiment orientation analysis on word pairs according to the extracted sentiment subjects and sentiment words as well as introduced sentence structure features. According to the invention, the sentiment subjects and the sentiment words in review sentences are extracted together, further, sentiment classification accuracy in the sentiment orientation analysis is improved, so that the sentiment element extraction and sentiment judgment are improved, and F value is up to 76.3%; due to introduction of word meaning codes, the generalization ability and the robustness of a system are improved by virtue of the word meaning codes, and the accurate rate and recall rate of review result analysis are greatly improved.

The invention relates to the technical field of log audit and safety management, and aims at providing a word segmentation algorithm-based log parsing method and a word segmentation algorithm-based log parsing system. The word segmentation algorithm-based log parsing method comprises the following steps: performing segmentation on a log, performing word sense analysis on segmentation results, performing word sense filtration on obtained segmentation results with word sense tagging, performing feature extraction on the obtained filtered segmentation results with the word sense tagging, performing feature matching on obtained word sense order feature codes, and performing semantic parsing on obtained semantic parsing rules; the word segmentation algorithm-based log parsing system comprises a segmentation module, a word sense analysis module, a word sense filtration module, a word order feature extraction module, a feature matching module and a semantic parsing module. According to the word segmentation algorithm-based log parsing method and the word segmentation algorithm-based log parsing system disclosed by the invention, the difficulty and complexity of log parsing are greatly reduced, and therefore the efficiency of performing parsing rule development on the log is increased; the word segmentation algorithm-based log parsing method and the word segmentation algorithm-based log parsing system can be better adapted to certain changes of a log format.

The invention provides a relation extraction method based on Bi-LSTM input information enhancement and belongs to the field of artificial intelligence natural language processing of computers. The method comprises the steps that by applying a strategy annotation dataset of an indeterminate label, a redundancy encoding technology is used for conducting character-level encoding on each word to generate a word form encoding vector; the word form encoding vector and a word embedding vector are spliced to generate a word vector used for capturing word form and word meaning information; Bi-LSTM of input information enhancement is used as a model encoding layer, the word vector is input to an encoding layer, and the encoding vector is output; the encoding vector is input into a decoding layer, and a decoding vector is obtained; by applying three layers of NN, an entity label, a relation type and entity number information are extracted from the decoding vector; finally, the gradient is calculated, the weight is updated, and a model is trained through a maximum target function. By means of the relation extraction method, the robustness of the system is improved, interference information caused by non-entity words is reduced, and the accuracy rate and recall rate of relation extraction are effectively increased.

Embodiments of the invention provide a word vector learning model training method and a server. The method comprises the following steps of: obtaining a word meaning vector and a primitive vector corresponding to a word in a training sample text; expressing a word vector in the training sample text according to the word meaning vector or the primitive vector; and substituting the word vector expressed by the word meaning vector or the primitive vector into an existing word vector learning model and carrying out model training. The server is used for executing above method. According to the word vector learning model training method and the server provided by the invention, the word vector is expressed through the primitive vector or the word meaning vector, and the word vector learning model corresponding to the word vector expressed by the word meaning vector or the primitive vector is trained, so that the semantic information of the word can be sufficiently expressed.

The invention provides a multi-subject extraction method based on a concept vector model. The method includes the following steps that firstly, a document is preprocessed through a traditional method and then vectors formed by feature words are preliminarily acquired; then synonyms are merged through the corresponding relation between word meanings and concepts in Hownet, disambiguation is conducted on polysemes through correlation between semantic classes and contexts, and the concept vector model is established to represent the document; concept similarity is calculated through related semantic information of the concepts in Hownet, a K-means algorithm is improved through a 'preset seed' method for clustering of the concepts, and then a plurality of subject concept clusters are formed; eventually, according to the corresponding relation between the concepts and words, a plurality of sub subject term sets are acquired. According to the method, semantic information is taken into consideration, the defects of sensitivity of the K-means algorithm to an initial center, space-time cost instability and the like are overcome, and the quality of extracted subjects is improved.

The embodiment of the invention provides an answer determination method and device for intelligent questions and answers. The method comprises the steps of taking an association statement set containing the current questions and the original dialogue context related to the current questions as a basic analysis object; introducing the context feature vectors of the associated statement set and performing answer prediction based on the context feature vector so as to distinguishing the word meaning of the same word in a specific context without needing to predefine various context scenes; meanwhile, using a trained context understanding model for carrying out feature vector interaction processing on the selected candidate answers and all associated statements in the associated statement set;based on the obtained multiple vector interaction matrixes, and identifying the question and answer matching degree of the candidate answers and the current question, so that the matching degree of the associated statement set and the candidate answers in the corresponding context can be accurately understood, the accuracy of selecting the target answers from the candidate answers is improved, and further the answer recommendation accuracy for the user consulting questions is improved, and the use experience of the intelligent question and answer user is improved.

The invention discloses a word-meaning disambiguation method by using a computer, a relevant system and searching method. The invention makes use of contribution degrees of background characteristic words of an ambiguous word to each meaning item to define the meaning of the ambiguous word and the computer can automatically learn disambiguation. The disambiguation realizing method of the invention is simple and high effective and can be understood easily. The experiment demonstrates that an average accuracy of the word meaning disambiguation carried out by the system can reach more than 90 percent. Simultaneously, compared with other disambiguation methods, the method can be operated more quickly and be realized more easily by the system. And the searching method adopted by the invention just shows a user the result which is identical to the meaning of the key word inputted by the user so as to largely enhance the searching efficiency.

The invention discloses a word multi-prototype vector representation and word sense disambiguation method based on CRP clustering, which comprises the following steps: the text in the massive text corpus is purified and pretreated to obtain plain text, CRP algorithm is used to cluster the context window representation of target polysemous word in the text corpus set. The target polysemous words inthe text corpus set are marked according to the clustering classification, and the polysemous words are trained on the marked text corpus set to obtain the multi-prototype vector representation of the polysemous words; 2, the target short text is preprocessed to obtain a short text word sequence, a target polysemous word in a word sequence is identifued, the contextual window of the target polysemous words is used to represent the similarity between the centroids of clusters corresponding to the words in the text corpus, and the word vector corresponding to the maximum similarity clusters isused as the word vector representation of the specific meaning of the polysemous words in the context to disambiguate the meanings of the polysemous words. The invention solves the problem of polysemyexpression in word expression and the problem of ambiguity identification in word meaning expression.

The invention belongs to the technical field of information security, and discloses a searchable encryption method and system for supporting fuzzy retrieval of word morphology and word meaning in cloud environments. The method includes establishing a security index by adopting a Bloom filter structure to support file dynamic update; accomplishing word morphology fuzzy matching by adopting a locally sensitive Hash function; performing word meaning expansion on keywords by using a natural language toolbox, constructing a word meaning fuzzy set, and inserting the word meaning fuzzy set into the established security index so that support word meaning keyword retrieval can be converted into accurate matching retrieval; performing inner product operation to realize the ordering of retrieval results by establishing query vectors and index vectors; and using the Hash function and a security knn algorithm to make the method and the system effectively resist selection keyword attacks so that thesearching of cryptograph on a cloud server under semi-trusted conditions. Cryptograph searching functions can be achieved; and the method and the system can support fuzzy retrieval of word morphologyand word meaning, multi-keyword retrieval, and dynamic adding and deleting files.

The invention relates to a Vietnamese news event detection method fusing dependency syntax information and a convolutional neural network, and belongs to the technical field of natural language processing. The method comprises the following steps: firstly, collecting a Chinese-Vietnamese bilingual news text, and setting an event type and a labeling system for event detection according to the characteristics of an event to form training data; and then fusing a convolutional neural network depending on syntax information, and detecting a sentence-level Vietnamese news event. Firstly, word meaning, position information, part-of-speech information and named entity information are fused in the encoding process. Secondly, two parts of features as event codes are fused by utilizing features between consecutive words of traditional convolutional coding and features between consecutive words of convolutional coding fused with dependency syntax information, so as to realize news event detection.The method has a good effect in news event detection.

The invention provides a terms disambiguation method based on a semantic dictionary. In the method, term concept relevance is calculated to realize pretreatment on automatic text summarization, namely term disambiguation; the correlativity of concepts is used for realizing the term disambiguation method; requirements of the concept, concept paraphrase, synonym of the concept, expanded paraphrase of the concept, synset of expanded concept and other factors as well as sentence coherence are comprehensively considered; and a correlation degree calculation formula of the concept and a backtracking method are used for selecting the optimum meaning of a word, thus realizing the semantic disambiguation based on the context. An experiment proves that recall and accuracy of the semantic disambiguation can be improved and the method can be better used in acquisition of text summarization.

The invention discloses an emotion analysis method and device, and related to the technical filed of natural language processing. The accuracy rate of the emotion analysis can be improved through the emotion analysis method and device. According to the technical scheme, the emotion analysis method comprises the steps of obtaining a semantic feature vector and a word meaning feature vector of a target text; conducting mixing of the semantic feature vector and the word meaning feature vector of the target text to obtain a target feature vector; conducting classification on the target feature vector according to a preset classification model to obtain a corresponding emotion mark, wherein a corresponding relationship between the target feature vector and the emotion mark is stored in the preset classification model; adopting the obtained emotion mark as an emotion analysis result of the target text. The emotion analysis method and device are mainly used for analyzing the emotion of the target text.

The invention relates to a word sense disambiguation method and device based on a word vector. The method comprises the following steps that: data preprocessing: carrying out processing, including punctuation removal, word segmentation and the like, on a document and a semantic item; training the word vector: using a word vector training tool to train the word vector; carrying out context vector representation: obtaining the word vector, and adopting a local weighting method to calculate the context vector; carrying out semantic item vector representation: obtaining the word vector of each word of the semantic item, and carrying out calculation to obtain a semantic item vector; carrying out similarity calculation: calculating a cosine similarity between the context vector and each semanticitem vector; carrying out semantic item distribution frequency calculation: carrying out statistics on the distribution frequency of each semantic item of an ambiguous term in a dataset; and carryingout final score statistics: calculating the cosine similarity between the context and each semantic item and the comprehensive score of each piece of semantic item frequency, wherein the semantic item with a highest score is an optimal word meaning.