107results about How to "Disambiguation" patented technology

The intelligent automated assistant system engages with the user in an integrated, conversational manner using natural language dialog, and invokes external services when appropriate to obtain information or perform various actions. The system can be implemented using any of a number of different platforms, such as the Web, email, smartphone, and the like, or any combination thereof. In one embodiment, the system is based on sets of interrelated domains and tasks, and employs additional functionally powered by external services with which the system can interact.

The invention discloses a method and system for news event extraction based on neural network. The method comprises the steps of: conducting data pre-processing on original text of training corpus; introducing an event sentence sequence represented by a word vector into a bidirectional long and short memory network, using the bidirectional long and short memory network to train and obtain semantic features of each candidate trigger word; introducing the event sentence sequence represented by a word vector into a convolutional neural network, using the convolutional neural network to train and obtain global features of the event sentences where the candidate trigger words are in; according to the semantic features of candidate trigger words and the global features of the event sentences where the candidate trigger words are in, using softmax as a classifier to classify each candidate trigger word, and therefore finding out the trigger words for news events, and according to the trigger word type, judging the type of the event. The method and system can quickly and accurately extract news events and deal with news events contained in non-standard statements, and has the advantages of high efficiency and universal applicability.

The invention provides an individuation searching method based on book domain ontology, belonging to an individuation network searching service. The individuation searching method comprises the following steps of: establishing the domain ontology, introducing a collaborative filtering idea, and adding semantic relationships which present influences among users; analyzing and processing logs and establishing a user model based on user interests and preferences; computing individuation scoring through spreading Activation (SA) based on the user model and the domain ontology; in addition, rearranging searched results, rearranging the results returned by a primary search engine according to an individuation scoring sequence from high to low, and returning to the users. Through introducing the collaborative filtering idea into the domain ontology, establishing the user model which presents user interest changes timely, and analyzing user requirements accurately by the SA, the invention effectively eradicates ambiguity of key words and greatly improves satisfaction degree of the users on the searched results.

The invention discloses a real-time intelligent monitoring method based on natural video. The method uses the knowledge of computer image processing and artificial intelligence and realizes unmanned intelligent monitoring and alarm to the action of pedestrian in public places and important sensitive places. Firstly, video frame serial sections which need to be studied are extracted, movable historical images are obtained, which reflect the movement process of the people; on this base, the user-defined method for extracting the eigenvector is used, vector representation of the specific movement series is obtained, and the vector sample is stored in the sample database; as for the video frame serial which need to be monitored, the eigenvector and sample data are mapped in the low dimensional space, the corresponding classification by the optimized method is obtained and alarm is carried out. Owing to the sample study mechanism and the classification mechanism of the designed actions in the text, the method of the invention improves the accuracy of identification and strengthens the expansibility of identification; by designing the eigenvector representation and extraction method of movement serial of the people, the completeness and accuracy of action representation are strengthened.

The invention discloses a novel method, software and a system about information intelligent retrieval, mining, filtering, graphic display and retrieval automation. The invention can carry out artificial intelligence information query, information monitoring and carry out searching, filtering, graph displaying and tracing according to the concept, so as to assist the user to carry out intelligent, high- quality searching and mining to especially large amount of information data in the internet or in a local computer. The method of the invention can extract important concepts and creative concepts contained on web pages and documents, ordering the concepts and displaying relations between the concepts by graphs. The product formed by the invention can be applied to such fields as the enterprise management and planning, the consumption market, the market research, the scientific research, the technology development, the middle-high education, the military affairs, the nation security and the diplomacy.

The invention provides a three-dimensional gesture action recognition method based on depth images. The three-dimensional gesture action recognition method comprises the steps of acquiring the depth images including gesture actions; dividing a human body region corresponding to the gesture actions from the images through tracking and positioning based on quick template tracking and oblique plane matching to obtain a depth image sequence after the background is removed; extracting useful frames of the gesture actions according to the depth images after the background is removed; calculating three-view drawing movement historical images of the gesture actions in the front-view, top-view and side-view projection directions according to the extracted useful frames; extracting direction gradient histogram features corresponding to the three-view drawing movement historical images; calculating relevance of combination features of the obtained gesture actions and gesture action templates stored in a pre-defined gesture action library; using a template with largest relevance as a recognition result of a current gesture action. Therefore, three-dimensional gesture action recognition can be achieved by adopting the three-dimensional gesture action recognition method, and the three-dimensional gesture action recognition method can be applied to recognition of the movement process of simple objects.

The invention provides a system and a method for detecting a surface defect of a glass panel. According to the system and the method, a single lens and a telecentric light path are combined with phase measurement deflectometry. Through setting positions of coding surface structural light display equipment, a beam splitter mirror, a to-be-detected surface of the glass panel and an image acquisition device in a space, a reflection law is satisfied. A lens is arranged between the coding surface structural light display equipment and the beam splitter mirror. Furthermore the coding surface structural light display equipment is arranged above an object space focal plane of the lens. Through computer controlling, standard N-step phase shift surface structural light which is perpendicular with each other is generated and transmitted to the to-be-detected surface of the glass panel. After image acquisition, back-end image processing equipment performs processing for obtaining modulation degree distribution, optimal gradient distribution and optimal height distribution of deformation surface structural light. The system and the method settle a problem of phase and height information inaccuracy caused by system ambiguity in defect detection by a traditional system based on the phase measurement deflectometry, and furthermore improving precision and accuracy in performing surface defect detection on the glass panel.

The invention discloses a domain knowledge graph based on a semantic triangle and a query method. The graph comprises a concept layer and an entity layer; the concept layer is composed of a set of concepts, and each concept has a unique identification and is represented by a represent word item and a candidate word item set; according to field related knowledge and the concept set of the concept layer, multiple entities corresponding to each concept are instantiated, the entity layer is formed, the entities are extensions of concepts, and each entity has a unique identification and is represented by a represent word item and a candidate word item set; an association relationship established according to the field related knowledge is set between the related concepts; an association relationship established according to the field related knowledge is set between the concept layer and the entity layer; an association relationship established according to the field related knowledge is set between the entities. By means of the method, separation of the concepts and the entities is achieved, knowledge summarization is convenient, different effects of the concepts and the entities in knowledge understanding and application are distinguished, and the query efficiency is improved.

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 utility model discloses a measurement method for same frequency and pilot frequency in a multi-carrier wave wireless communication system, comprising the following steps: a new measure series are added in the measure control information, in the measure series, a plurality of measure cells are identified by the main carrier frequency and the Cell Parameter ID. The measure series comprises a P-CCPCH RSCP measurement entity and a ISCP measurement entity of a user terminal based on the neighborhood relation. The utility model has advantages of providing a measurement method for a multi-carrier frequency system serving cell and neighborhood P-CCPCH RSCP and ISCP, eliminating the polysemy of the prior same frequency or pilot frequency, meanwhile furthest being compatible with the content of 3GPP and CCSA on protocol and concept definition.

The invention provides a natural-language-processing method of an ancient-tablature and ancient-culture knowledge graph. The method includes: creating patterns of the knowledge graph of ancient tablature and ancient culture according to ancient-tablature and ancient-culture expertise; acquiring data, and extracting information; fusing knowledge, wherein the step includes linking entities and merging the knowledge; and integrating ancient-tablature and ancient-culture knowledge library patterns and data to solve conflicts of the patterns and data. According to the method, natural-language-processing technology and knowledge graph technology are used in the field of ancient-tablature research, graphics are used to show development processes and structure relationships of ancient-tablature knowledge, visualization technology is used to describe ancient-tablature knowledge resources and carriers thereof, and the shown knowledge and interrelationships among the same are mined, analyzed, constructed and drawn. A core structure, development history and an overall architecture can be demonstrated, multidisciplinary fusing can be achieved, and valuable references can be provided for ancient-tablature discipline research.

The invention provides a WSDL semi-structured document similarity analyzing and classifying method based on a semantic model. The method includes the steps that a WordNet dictionary is used for establishing a WSDL semi-structured document semantic model, lexical ambiguity is eliminated through a maximum entropy model, a WSDL semi-structured document corpus feature vector model is established, a document feature matrix of WSDL semi-structured documents is generated, hence, content classification and evaluation are conducted on two different documents, and finally the similarity comparison of service functions is obtained. By means of the WSDL semi-structured document similarity analyzing and classifying method based on the semantic model, the judging accuracy of document similarity is improved, the document classification speed is increased, the document classification precision is improved, and a dimensionality reduction effect can be achieved on vector space.

The invention discloses a method and a device for constructing a power grid fault processing knowledge graph. Domain high-quality professional phrase mining and named entity identification are carriedout; relationship extraction of equipment, events, operations and the like is carried out; knowledge is automatically extracted from massive structured and unstructured data such as power grid operation data and fault processing plans; a power grid dispatching fault processing knowledge graph is constructed, the limitation that a traditional power grid dispatching expert system customizes knowledge regularly and is long in period and slow in updating is well broken through, and computer intelligent autonomous collection, cross-domain analysis and all-time learning of the power grid dispatching fault processing knowledge are achieved.

The invention discloses a power grid tripping event intelligent identification method based on a knowledge graph. The method is based on a knowledge graph technology and a natural language processingtechnology in the field of artificial intelligence. Based on power system knowledge characteristics, the monitoring service work flow and the power grid tripping event judgment logic are deeply studied; based on learning of historical sample data of a power grid tripping event and through knowledge extraction and fusion, the method constructs a power grid tripping event knowledge triple model, forms an association relationship among the entities and completes disambiguation of information with uncertainty, thereby reducing understanding difference caused by improper personnel maintenance or professional barriers and realizing intelligent identification of power grid tripping events; knowledge reasoning is performed based on the knowledge graph, cross confirmation and more information collection are completed in combination with related power grid state perception, and the response efficiency of power grid tripping events is improved while the service pressure of power grid monitoring personnel is effectively reduced.

The invention relates to the hot rolling strip steel rolling technology, in particular to the virtual simulation of hot continuous rolling strip steel rolling process, and discloses a hot continuous rolling strip steel virtual rolling simulation system which solves the problems during strip steel rolling including equipment problem, defective workmanship, system logic loophole and model precision by means of simulation technique. The hot continuous rolling strip steel virtual rolling simulation method comprises the following steps that: simulation equipment and a simulation input control strategy triggering system used during hot continuous rolling are set up; operating condition is input to the simulation equipment by the simulation input control strategy triggering system; the simulation equipment operates to obtain equipment data and basic automation level data; the equipment data and basic automation level data are converted and transmitted to real equipment, equipment after virtual execution conversion and the basic automation level data; moreover, execution results are monitored. The invention has the advantages that: the invention substantially reduces on-site maintenance and debugging time and greatly improves control quality and precision; moreover, the invention provides a visualized communication platform between designers and end users.

The invention provides a data processing method based on a data warehouse. The method comprises the following steps: an atomic layer acquires the structured first data; the atomic layer performs organization and classification on the first data according to a preset way so as to acquire the second data; the atomic layer performs partitional storage on the organized and classified second data; theatomic layer updates the second data to an integration layer; the integration layer performs merging processing on the second data according to a preset integration rule so as to generate the third data to store. The data inconsistency caused by data redundancy, data missing and like circumstances can be effectively avoided since the atomic layer performs organization and classification on the first data according to the preset way, and the maintenance cost is reduced; the partitional storage is performed on the organized and classified first data so that the information can be conveniently traced; the integration layer performs the merging processing on the second data so that the multi-source data can be integrated as the unique information, and then the ambiguity of the data is eliminated, and the unique accuracy of the data is guaranteed.

The invention relates to a neural machine translation method based on Multi-BiRNN encoding. Multi-BiRNN encoding is adopted at an encoder end, that is to say that on the basis of using source languagesentences as input sequences, one or more groups of BiRNN are added to encode other input sequences associated with the input sequences; on the basis of Multi-BiRNN-encoded neural machine translation, in the encoding process of a source end, the source language sentence sequences and a dependency syntax tree thereof are considered at the same time, and serialization results of the syntax tree areobtained by means of two different traversal modes respectively and serve as Multi-BiRNN encoding input with the source language sentence sequences; at the output end of each BiRNN, a word is formedin a vector splicing mode. According to the method, vectors obtained by encoding contain more abundant semantic information, the source language sentence sequences and other sequences associated withthe source language sentence sequences are considered at the same time, and the disambiguation function is achieved in the semantic representation process of the source language sentences.

The invention relates to a method for coarsely aligning a static base based on polarization and gravity information. Adopted equipment comprises a tri-axial accelerometer, a GPS receiver and two or more polarization sensors. The method comprises the following steps: firstly, calculating a unit sun vector under a geographic coordinate system by utilizing geographic position information of a carrier and time information; secondly, obtaining gravity information by utilizing the accelerometer; thirdly, obtaining sun vector information under a carrier coordinate system by utilizing the polarization sensors; finally, determining a vector posture transformation matrix by utilizing the information. The method provided by the invention has the advantages of being high in precision, low in calculated amount and the like.

The invention discloses a BERT-based multi-feature fusion fuzzy text classification model. The method comprises the following steps of preparing a fuzzy text classification original data set; a BERTMFFM model is constructed, the BERTMFFM model comprises a BERT model, a convolutional neural network, a bidirectional long-short memory network and a SelfAttention module, the input of the BERT model is a fuzzy text, the output of the BERT model is connected with the convolutional neural network, the bidirectional long-short memory network and the SelfAttention module, and local features, sentence semantic features and syntax structure features of the fuzzy text are extracted; splicing the output of the BERT model with the output of the bidirectional long-short memory network at the same time, and then screening out optimal sentence semantic features by using maximum pooling operation; and fusing the local features, the optimal sentence semantic features and the syntactic structure features by adopting a parallel splicing mode, and performing fuzzy text classification on a fusion result through a SoftMax function to finish the construction of the BERTMFFM model. The problem of incomplete feature collection is solved, so that the classification accuracy is improved.

The invention provides a bilingual sentiment classification method and device. The method comprises the steps of translating original language documents to be classified and the original language documents of a training sample set to obtain the translated documents to be classified and the translated documents of the training sample set; combining the original language documents to be classified and the translated documents to be classified to obtain a bilingual documents to be classified, and combining the original language documents of the training sample set and the translated documents of the training sample set to obtain bilingual documents of the training sample set; establishing a bilingual feature vector space to be classified and a bilingual feature vector space of the training sample set; training classifiers on the bilingual feature vector space of the training sample set by using a maximum entropy model; carrying out sentiment polarity classification on the bilingual feature vector space to be classified through the trained classifiers. The bilingual sentiment classification method and device combine the characteristics of two languages, provide extra classification information for the sentiment classification, improve classification accuracy, extract important characteristic items from the bilingual feature vector spaces, and improve classification efficiency.

The invention provides an auxiliary control method and system based on post verification. The method comprises the steps that through the pre-processing of original command signals, obtaining auxiliary control information, conducting grammatical template matching of the auxiliary control information based on fields to obtain command operator parts and operation parameter parts, using the post verification method corresponding to the fields to analyze the operation parameter parts, treating the result which is obtained from the analysis and with the highest matching level as the operation parameters of the command operator parts, generating auxiliary control commands and sending the control commands to a central processing unit for operation, thus achieving the auxiliary control. The invention hugely enhances the precision of semantic identification based on current grammatical retrieval. By simulating the manner of human understanding of language, all the semantic meanings which can be extracted from the colloquial expression content inputted by the user are identified, and the identification rate and the user's experience are enhanced.

The invention discloses an information processing method and device. The method comprises the following steps that judgment is made whether an ambiguous term exists in a text to be processed; when the ambiguous term exists in the text to be processed exists, the ambiguous term is split from the text to be processed; the split ambiguous term is split according to a term splitting rule corresponding to the split ambiguous term. According to the information processing method and device, ambiguous term judgment is conducted on the text to be processed, when the ambiguous term exists in the text to be processed, the ambiguous term is split form the text to be processed, and the split ambiguous term is split according to the term splitting rule corresponding to the ambiguous term. The information processing method and device effectively remove the ambiguous term in the text to be processed, and the accuracy rate for information processing is improved.

A method is disclosed for quantitatively assessing information in natural language contents related to an object name. The method includes identifying a sentence in a document, determining a subject and a predicate in the sentence, and retrieving an object-specific data set related to the object name. The object-specific data set includes property names and association-strength values. Each property name is associated with an association-strength value. The method also includes identifying a first property name in the property names that matches the subject, assigning a first association-strength value associated with the first property name to the subject, identifying a second property name in the property names that matches the predicate, assigning a second association-strength value associated with the second property name to the predicate, and multiplying the first association-strength value and the second association-strength value to produce a sentence information index.

The invention provides a chronic non-infectious disease risk prediction system based on a knowledge graph. Based on the characteristics that chronic non-infectious diseases are not easy to find in the early stage and have long onset time, and patients are rejuvenated, the advantage that the knowledge graph can quickly and effectively obtain related knowledge and the logical relationship between the knowledge is utilized, the knowledge graph is constructed by utilizing the core technologies such as a decision tree algorithm and a BP neural network algorithm, and a disease risk prediction result is more conveniently and accurately provided for a user, so that the pressure of a medical system is reduced.

The embodiment of the invention discloses a text processing method and device. The text processing method and device are used for conveniently classifying text data. The text processing method comprises the steps of obtaining a text to be classified, and conducting word splitting on the text to be classified to obtain a word splitting result; according to the word splitting result, building a target characteristic vector; using a pre-established SVM classifier to analyze the target characteristic vector to obtain a target label, wherein the SVM classifier is built according to at least two kinds of corresponding relations of the characteristic vector and the label, the characteristic vector is obtained by building text information, and the labels of different kinds of corresponding relations are different. Since when the SVM classifier is built, the text information is utilized, thus the text to be classified can be classified, the classifying result is the target label, that is, the SVM classifier outputs the target label to be used for identifying the text to be classified, and therefore, through the target label, the text to be classified obtains a sole identification.

The invention relates to the technical field of natural language processing, and provides a text matching method. The method comprises the following steps: preprocessing the text to be matched, wherein the preprocessing comprises word segmentation processing of the text to be matched; splicing the mergeable participles into phrases; calculating a similarity between the text to be matched and the pre-stored text in a text library based on a first similarity algorithm, and selecting a preset number of the pre-stored text meeting preset conditions as the first similar text; calculating a similarity between the text to be matched and the first similar text based on a second similarity algorithm; selecting a preset text whose similarity satisfies a preset condition from the first similar text,and determining the selected text to be a matching text. The text matching method provided by the invention can improve the accuracy of the text matching.

The invention relates to a Vietnamese news event detection method based on Chinese information and Vietnamese statement method guidance, and belongs to the technical field of natural language processing. The method comprises the following steps: firstly, enabling Vietnamese sentences to obtain Chinese semantic information in a bilingual information fusion module; then, according to the Vietnamese statement dependency relationship, utilizing syntax graph convolution to enhance the association between information, so that the problem of multi-event ambiguity is solved; and finally, extracting Vietnamese sentence representation based on the Chinese event type by utilizing an event type perception network so as to complete Vietnamese event detection. Event detection needs to identify event information in news and classify the event information, at present; aiming at the problems of scarcity of Vietnamese news event annotation data and ambiguity of event types, according to the multi-language consistency characteristic, rich Chinese annotation data can be used for solving the problems of scarcity of Vietnamese data and ambiguity of event types.

The invention discloses a method, device and equipment for determining a search result and a computer storage medium, and relates to the field of intelligent search. According to the specific implementation scheme, the method comprises the steps that 1, obtaining the current query of the user, the search history information of the user in the first duration, the search history information of the user in the second duration and the candidate search result of the current query; a search result ranking model is input, a search result corresponding to the current query is determined according to scores of the candidate search results, and the second duration is longer than the first duration; wherein the score of the candidate results by the search result sorting model is determined accordingto the similarity between the integration of the vector representation of the current query and the vector representation of the search history information of the user within the first duration and the vector representation of the candidate search results; and determining the similarity between the integration of the vector representation of the current query and the vector representation of the search history information of the user within the second duration and the vector representation of the candidate search result.

The invention relates to a deep learning commodity comment text sentiment tendency analysis method. The method comprises the following steps of processing the text to obtain the simplest text; dividing the simplest text into a training set and a test set; obtaining a word sequence according to the training set and setting an emotion feature with an emotion weight value, adjusting the emotion weight value to obtain a corrected emotion weight value; according to the word sequence, obtaining a plurality of word group sequences according to different composition word combination modes; calculatinga correction emotion weight value in the word group sequence to obtain an emotion label of the text; and comparing the emotion label with the human emotion evaluation to obtain a final emotion weightvalue and a final corrected emotion weight value, executing the operation on all texts in the training set, and obtaining an emotion model according to the final emotion weight value and the final corrected emotion weight value to verify the emotion analysis of the test set.