516 results about "Semantic relation" patented technology

A process and system for database storage and retrieval are described along with methods for obtaining semantic profiles from a training text corpus, i.e., text of known relevance, a method for using the training to guide context-relevant document retrieval, and a method for limiting the range of documents that need to be searched after a query. A neural network is used to extract semantic profiles from text corpus. A new set of documents, such as world wide web pages obtained from the Internet, is then submitted for processing to the same neural network, which computes a semantic profile representation for these pages using the semantic relations learned from profiling the training documents. These semantic profiles are then organized into clusters in order to minimize the time required to answer a query. When a user queries the database, i.e., the set of documents, his or her query is similarly transformed into a semantic profile and compared with the semantic profiles of each cluster of documents. The query profile is then compared with each of the documents in that cluster. Documents with the closest weighted match to the query are returned as search results.

A speech recognition apparatus includes a first-candidate selecting unit that selects a recognition result of a first speech from first recognition candidates based on likelihood of the first recognition candidates; a second-candidate selecting unit that extracts recognition candidates of a object word contained in the first speech and recognition candidates of a clue word from second recognition candidates, acquires the relevance ratio associated with the semantic relation between the extracted recognition candidates of the object word and the extracted recognition candidates of the clue word, and selects a recognition result of the second speech based on the acquired relevance ratio; a correction-portion identifying unit that identifies a portion corresponding to the object word in the first speech; and a correcting unit that corrects the word on identified portion.

The invention provides a text emotion classification method based on the joint deep learning model which relates to the text emotion classification method. The method is designed with the object of solving the problems with the dimension disaster and sparse data incurred from the existing support vector machine and other shallow layer classification methods. The method comprises: 1) processing each word in the text data; using the word2vec tool to train each processed word in the text data so as to obtain a word vector dictionary; 2) obtaining the matrix M of each sentence; training the matrix M by the LSTM layer and converting it into vector with fixed dimensions; improving the input layer; generating d-dimensional h word vectors with context semantic relations; 3) using a CNN as a trainable characteristic detector to extract characteristics from the d-dimensional h word vectors with context semantic relations; and 4) connecting the extracted characteristics in order; outputting to obtain the probability of each classification wherein the classification with the maximal probability value is the predicated classification. The invention is applied to the natural language processing field.

A structural natural language index is created by segmenting documents within a repository into text portions and extracting named entity, co-reference, lexical entries, structural-semantic relationships, speaker attribution and meronymic derived features. A constituent structure is determined that contains the constituent elements and ordering information sufficient to reconstruct the text portion. A functional structure of the text portions is determined. A set of characterizing predicative triples are formed from the functional structure by applying linearization transfer rules. The constituent structure, the characterizing predicative triples and the derived features are combined to form a canonical form of the text portion. Each canonical form is added to the structural natural language index. A retrieved question is classified to determine question type and a corresponding canonical form for the question is generated. The entries in the structural natural language index are searched for entries matching the canonical form of the question and relevant to the question type. The characterizing predicative triples are used in conjunction with a generation grammar to create an answer. If the generation fails, some or all of the constituent structure of the matching entry is returned as the answer.

The invention provides an implement method of an FAQ Chinese question answering system in the tourism field. The implement method comprises the steps of FAQ collection and organization, construction of a tourism field knowledge base, user query, question analysis, answer extraction and the like, thereby realizing the FAQ Chinese question answering system in the tourism field. The implement method constructs the tourism field knowledge base-field knowledge network with the help of the idea of ontology, utilizes the KDML language to define and describe the terms and the relations of the tourism field and realizes the integration of the tourism field knowledge network and a general knowledge base-knowledge network. The invention proposes a calculation method of similarity of tourism questions on the basis, the method realizes the calculation of the similarity of the questions with the help of the characteristics of the questions of the tourism field and the combination of morphological relations, the syntactic dependency relations and the field concept semantic relations in the questions, and the method further searches the related question from a candidate question set and extracts the answer of the question based on the similarity calculation. The test result of a Yunnan tourism FAQ question answering system proves that the method is feasible and has better effect.

A multi-modal natural language question answering system and method comprises receiving a question logic form, at least one answer logic form, and utilizing semantic relations, contextual information, and adaptable logic.

The present invention is to provide a learning machine capable of expressing / accumulating concept and semantic relation by understanding semantic relation of information as a relation between concepts after characteristics, semantic relation and structure of information and the like have been analyzed. The present invention is intended to realize a thinking machine in which information is inclusively collected and stored as knowledge structured based on the semantic relation of the collected information, information is generated by predetermined inference so as to have a new semantic relation in response to an inquiry or request and which can decide an optimal solution to an inquiry and the like by evaluating / deciding new generated information.

The invention discloses a semantic query expansion method based on domain knowledge, which comprises the following steps: taking concept expression and a knowledge tree system as the basis to construct the domain knowledge; performing primary semantic analysis on query phases input by users to form a semantic item list; utilizing results of the primary semantic analysis and taking the domain knowledge as the basis to construct a semantic map with expansion types and expansion weights; respectively computing semantic distances between each vertex and an initial vertex in the semantic map; determining an expandable item of each item in the semantic item list according to the semantic distances; and finally, combining all expandable items according to AND / OR logic relations to obtain a semantic item set representing the query intension of the users, and submitting the semantic item set to a searching system for searching. In the semantic query expansion method based on the domain knowledge, the computing time is short, the domain knowledge is fully utilized, and newly-added expanded semantic items and the original query phases have definite semantic relations, and the recall ratio and the precision ratio of the searching system can be improved effectively.

The embodiment of the invention provides an intelligent question-answer implementing method and system. The method comprises receiving a question to be queried; performing semantic analysis on the question; according to the result of the semantic analysis, performing corresponding retrieval treatment on the question, wherein the retrieval treatment comprises retrieval treatment on the question by at least one of a semantic relation digging system, a text library retrieving system, a knowledge base retrieving system and a question-answer library retrieving system; feeding an answer back according to the result of the retrieval treatment. The intelligent question-answer implementing method effectively improve the accuracy of question answering.

The invention discloses a deep learning-based blog text abstract generation method. The method comprises the following steps of: crawling blog data; preprocessing the crawled blog data and selecting blog text data; converting the selected blog text data into vector matrix data according to a Chinese word vector dictionary; constructing a deep learning encoder-decoder model, separately training an encoder and a decoder of the model, and connecting the encoder and the decoder for use after the training is completed; and repeating the steps S01 to S03 to obtain generated data, and generating a predicted abstract from the generated data through the trained model. According to the method, text abstracts of blogs are automatically generated on the basis of a deep learning frame encoder-decoder, and deeper semantic relation of the blogs can be obtained at the same time. The generated text abstracts can visually display the main content of the current blog, so that the text abstracts have wide application prospect.

A system and method of developing rules for text processing enable retrieval of instances of named entities in a predetermined semantic relation (such as the DATE and PLACE of an EVENT) by extracting patterns from text strings in which attested examples of named entities satisfying the semantic relation occur. The patterns are generalized to form rules which can be added to the existing rules of a syntactic parser and subsequently applied to text to find candidate instances of other named entities in the predetermined semantic relation.

A method or module for creating an Language Neutral Syntax (LNS) representation of a sentence from a language particular syntax representation such as found in the Penn Treebank for use by different applications. The method or module includes a node generator configured to create hierarchical and dependent nodes using phrasal and constituent nodes of the language particular syntax. A node dependency generator is configured to create an unordered hierarchical dependency structure for the hierarchical and dependent nodes using a semantic relation between the hierarchical and dependent nodes derived from the language particular syntax.

The invention discloses a neural network natural language inference method combining single word semantic knowledge. The method comprises the steps of extracting various semantic relations between word pairs from an external semantic knowledge base; constructing a neural network model, utilizing the neural network model to conduct local inference modeling on input prerequisite text, hypothetical text and the semantic relations between the word pairs, combining local inference information to obtain inference information at a sentence level, and training the neural network model at last; inputting the unlabeled prerequisite text and hypothetical text into the trained neural network model, calculating and obtaining probability distributions belonging to three categories, and selecting the category corresponding to the maximum probability as a final prediction category. The method improves the accuracy of natural language inference and solves the problem that external semantic knowledge isless utilized in traditional neural network schemes.

The invention provides a method for semantic relation extraction, wherein on the basis of an annotated training corpus having tokens with associated relational labels each indicating a relation between the respective token and a selectable key entity semantic relation between said key entity and other entities are directly extracted from unstructured text using a probabilistic extraction model.

A lexical knowledge base is compiled automatically from a machine-readable source (such as an on-line dictionary or unstructured text). The preferred embodiment of the invention makes use of “backward linking,” by which inverse semantic relations are discerned from the text and used to augment the knowledge base. By this arrangement, on-line dictionaries and other texts can provide formidable sources of “common sense” knowledge about the world.

The invention discloses an unmanned vehicle semantic map model building method and an application method thereof to an unmanned vehicle. Extraction of a conceptual structure indicates that key map elements such as road networks, road traffic participants and traffic rules related in the running process of the unmanned vehicle are reasonably abstracted into different conceptual types, establishment of the semantic relation between concepts refers to establishment of map concept semantic hierarchical relations and incidence relations, and living examples of the conceptual types and the semantic relation among the living examples are established in an instantiated manner to finally obtain a semantic map for the unmanned vehicle. A map data structure applicable to the unmanned vehicle is built, the sufficient semantic relation among the map elements is designed, the semantic map is generated, semantic reasoning is performed according to the semantic map, a globally planned route, the current position and orientation of the unmanned vehicle and peripheral real-time obstacle information to obtain local scene information of the unmanned vehicle, scene understanding of the unmanned vehicle is realized, and the unmanned vehicle is assisted in behavior decision.

A method for building a knowledge base containing entailment relations, includingproviding at least one input pattern (p) with N pattern slots (N>1), the input pattern (p) expressing a specific semantic relation between N entities that fill the N pattern slots of the input pattern (p) as slot fillers,providing at least one cluster (c) of articles, the articles of the cluster (c) relating to a common main topic;processing the articles with respect to the input pattern (p) and identifying the identities which match the semantic type of the N pattern slots;if the at least one input pattern matches a portion of an article (a) of the at least one cluster (c):storing the N slot fillers (s1, s2, . . . , sN), which match the slots of the pattern (p), and a cluster identifier Ic of the cluster (c) into a first table S, wherein the N-tuple (s1, s2, . . . , sN) and the cluster identifier Ic of the associated cluster (c) form one element of the table S;for each element of table S, identifying appearances of the slot fillers (s1, s2, . . . , sN) in a plurality of articles of cluster (c) and for each appearance so identified, storing the slot fillers (s1, s2, . . . , sN) together with the sentence in which they occur into a second table C0;from the sentences stored in table C0, extracting patterns which span over the corresponding N slot fillers (s1, s2, . . . , sN), the extracted pattern expressing a semantic relation between the N slot fillers; andstoring the extracted pattern together with the input pattern as entailment relation into the knowledge base.

The invention discloses a multilayer quotation recommendation method based on a literature content mapping knowledge domain, and belongs to the field of information recommendation and intelligent information processing. The method comprises the following steps: firstly, obtaining the query requirement of a user, wherein the query requirement consists of the key words of the title and the digest of a thesis which needs to recommend a quotation thesis or quotation literature; then, on the basis of the literature content mapping knowledge domain, expanding and querying a retrieval word, wherein the mapping knowledge domain consists of the research object word and the research behavior word node of the literature, and edges which express various semantic relations including synonymy, synonym, an up and down position, part-whole, juxtaposition and the like; and finally, constructing the inverted index of the literature in a data set, selecting a candidate quotation, calculating the similarity between the candidate quotation and query, and adopting a gradient progressive regression tree to carry out quotation recommendation. The method carries out multilayer quotation recommendation on the basis of the literature content mapping knowledge domain, enlarges the range of the candidate quotation, accurately expresses the research object and contents of the thesis, improves efficiency for users to obtain a relevant literature and has a wide application prospect.

A system and method that improves obtaining similarity measure between concepts based on Latent Semantic Analysis by taking onto account graph structure derived from the knowledge bases by using a vector propagation algorithm, in the context domain, such as a medical domain. Concepts contained in a corpus of documents are expressed in a graph wherein each node is a concept and edges between node express relation between concepts weighted by the number of semantic relations determined from the corpus. A vector of neighbors is created and assigned to each concept, thereby providing an improved similarity measure between documents, i.e., corpus and query against corpus.

The invention provides a method, an apparatus and a terminal unit for matching semantics with an image, wherein the method comprises: acquiring input sentences; carrying out semantic analysis of the sentences to acquire a plurality of feature words; according to the feature words, searching for image elements matching with the feature words in the image element database; according to the semantic relations among the plurality of feature words, determining the combination sequence among the image elements matching with the feature words; and according to the combination sequence, synthetizing the image elements matching with the feature words into a new image. By using the scheme of the invention, images matching with the semantics can be obtained, achieving an effect that images and accompanying texts are both excellent.

Compile-time non-concurrency analysis of parallel programs improves execution efficiency by detecting possible data race conditions within program barriers. Subroutines are modeled with control flow graphs and region trees having plural nodes related by edges that represent the hierarchical loop structure and construct relationship of statements. Phase partitioning of the control flow graph allows analysis of statement relationships with programming semantics, such as those of the OpenMP language, that define permitted operations and execution orders.

The present invention provides a system and method for automatically refining ontology within a specific context. The system comprises: a rich context extractor for discovering a semantics relationship conflict existing between an original ontology schema and application data; and an ontology schema and mapping refiner for refining the original ontology schema based on the semantics relationship conflict discovered by the rich context extractor, and creating a new mapping between the refined ontology schema and a data schema of the application data. According to the invention, users can save a lot of work of manual refinement, since the ontology schema is automatically refined based on contexts in the application data.

Provided are a system capable of providing a highlight of a real-time event using message streams of social media and a highlight service providing method using the same. The disclosed highlight providing system includes: a hot event session detection unit detecting a hot event session where a hot event occurs using message streams of social media; a keyword detection unit detecting keywords related with the hot event from the hot event session detected by the hot event session detection unit using a predetermined algorithm; a hot event generating unit generating the hot event using a semantic relation between the keywords detected by the keyword detection unit; and a highlight providing unit providing a highlight of the social media along with a description in a natural language based on the hot event generated by the hot event generating unit.

A method of intelligent textual markup in an information exchange includes: determining semantic elements in said information exchange; determining relations between said semantic elements; representing said semantic elements as nodes in a directed graph; and representing said relations as edges connecting said nodes. A data processing system for enabling a visual representation of semantic relations in an information exchange includes: a semantic analysis engine adapted to determine semantic elements of said information exchange; a relation analysis engine adapted to determine relations between said semantic elements; and a presentation engine adapted to present said semantic elements as nodes and said relations as edges in a directed graph representing said information exchange.

A system and method for classifying comments are disclosed. The method includes receiving a collection of comments. Each of the comments in the collection includes text in a natural language and is associated with a previously-submitted idea submission which includes a description of an idea. The method further includes natural language processing each of the comments to identify dependencies (syntactic and / or semantic relations between text elements) in at least a part of the comment. Based on the identified dependencies, the comments are each automatically classified into one (or more) of a plurality of comment classes. The comment classes may include a first class for reaction to the content of the idea, a second class for expression of a commenter's judgment of an idea's value, and a third class for reaction to an idea generation process in which the associated idea submission is made. Information based on the assigned comment classes is output. For example, actions are proposed to use the comments according to their class.

The invention relates to an advertisement content matching and effect evaluation technology, and provides a rich media internet advertisement content matching and effect evaluation method. The method comprises the following steps of: dynamically collecting network data, and processing multimedia content; analyzing the multimedia content, extracting features and establishing a feature library; establishing a feature management knowledge base and a semantic relation library based on an XML model; implementing semantic representation for the network advertisement content by a ontology base technology; implementing quick and accurate matching of text, voice, video, picture and the like through an Agent-based automatic semantic matching technology; generating a mass of record files and information bases from the representation of user elemental information, user behavior and user keywords through WEB data mining, thereby obtaining indexes, such as browse rate, click rate, paths, association rules, sequence modes, statistical characteristics and the like, and deeply mining and analyzing the indexes to evaluate the advertisement effect.

The invention discloses an automatic text summarization method based on enhanced semantics. The method comprises the following steps of: preprocessing a text, arranging words from high to low according to the word frequency information, and converting the words to id; using a single-layer bi-directional LSTM to encode the input sequence and extracting text information features; using a single-layer unidirectional LSTM to decode the encoded text semantic vector to obtain the hidden layer state; calculating a context vector to extract the information, most useful the current output, from the input sequence; after decoding, obtaining the probability distribution of the size of a word list, and adopting a strategy to select summarization words; in the training phase, fusing the semantic similarity between the generated summarization and the source text to calculate the loss, so as to improve the semantic similarity between the summarization and the source text. The invention utilizes the LSTM depth learning model to characterize the text, integrates the semantic relation of the context, enhances the semantic relation between the summarization and the source text, and generates the summarization which is more suitable for the subject idea of the text, and has a wide application prospect.

The invention relates to the text analysis field, particularly to a semantic characteristic-based text semantic similarity analysis method. According to the technical scheme, the similarity degree between texts is analyzed more accurately and effectively by calculation based on semantic relations of internal words of the texts. According to the method, shallow analysis on association relation between texts and between lexical items is performed through singular value decomposition; a lexical item-theme set is constructed by a bayesian network; the semantic similarity between the lexical items is calculated by mutual information and context; and finally, the text similarity is calculated through a graph structure. By adoption of the text semantic similarity analysis method, the semantic relation between texts can be measured and recognized more accurately and effectively.