Patents

Literature

Hiro is an intelligent assistant for R&D personnel, combined with Patent DNA, to facilitate innovative research.

5 results about "Named entity" patented technology

Filter

Efficacy Topic

Property

Owner

Technical Advancement

Application Domain

Technology Topic

Technology Field Word

Patent Country/Region

Patent Type

Patent Status

Application Year

Inventor

In information extraction, a named entity is a real-world object, such as persons, locations, organizations, products, etc., that can be denoted with a proper name. It can be abstract or have a physical existence. Examples of named entities include Barack Obama, New York City, Volkswagen Golf, or anything else that can be named. Named entities can simply be viewed as entity instances (e.g., New York City is an instance of a city).

Power operation and maintenance information knowledge graph construction method

InactiveCN108460136AImprove effective useImprove the level of intelligent applicationNatural language data processingSpecial data processing applicationsNamed-entity recognitionKnowledge extraction

The invention provides a power operation and maintenance information knowledge graph construction method. A deep learning algorithm is introduced in construction of a power field knowledge graph; andtwo major machine learning tasks of named entity identification and entity relationship extraction are adopted for solving two difficult problems of knowledge unit extraction and knowledge unit relationship extraction. The effective utilization of power data is improved; the hidden value of the data is exerted; the intelligent application level of a power network is improved; and based on knowledge representation of the knowledge graph, a low-dimensional and mutually isolated information display mode in the past is changed, and association among power knowledge is presented more intuitively. Apower information operation and maintenance system knowledge library is constructed; historical operation and maintenance data is converted into knowledge; online autonomous consultation and problemsolving of a user are supported; intelligent question and answer are realized; the operation and maintenance efficiency is improved; important values are exerted in power network operation and maintenance work; and based on this, a power information operation and maintenance field knowledge graph optimization storage technology is finished.

Owner:STATE GRID FUJIAN ELECTRIC POWER CO LTD +1

Text recognition method and device, electronic equipment and medium

PendingCN111723575ACharacter and pattern recognitionNatural language data processingText recognitionWord list

The invention discloses a text recognition method and device, electronic equipment and a medium. According to the invention, the method comprises the steps of: performing entity feature recognition onthe target text by using a pre-trained deep learning model to obtain a candidate name entity list, and matching the candidate name entity list with multiple pieces of name information in the enterprise name library one by one to obtain at least one matching result, thereby taking the candidate name entity, higher than the hit matching rate, in the at least one matching result as a name entity obtained by identifying the target text. By applying the technical scheme, name entities possibly existing in the text can be extracted by adopting the deep learning model, a part of entities with identification errors are filtered out by utilizing the filtering word list to serve as candidate company entities, and the candidate companies correspond to the specific enterprise name libraries by meansof the enterprise name libraries and the enterprise entity mapping tables. Therefore, the problem that the efficiency of extracting the effective name entity from the text is very low in the prior artis avoided.

Owner:HANGZHOU WEIMING XINKE TECH CO LTD +1

Internet of Things terminal interoperation method and device based on semantic comprehension

ActiveCN106933796AEfficient combinationImprove interoperabilitySemantic analysisTransmissionTask completionThe Internet

The invention discloses an Internet of Things terminal interoperation method and device based on semantic comprehension, and relates to the field of the Internet of Things and the Internet. The method comprises the following steps that: according to an internal knowledge base, an external knowledge base and Internet data contents, obtaining each entity oriented to an Internet of Things application and the name of each entity in the Internet; calculating the correlation of all entities, and forming a uniform named entity name which aims at the same entity; carrying out semantic extension on each entity and the operation function of each entity; when an Internet of Things request is received, through Internet of Things entity semantic calculation, determining the named entity name and an operation function corresponding to the Internet of Things request; when a certain operation is finished, recording and evaluating a current operation and task finishing situation; and according to the situation, carrying out back annotation on each entity which realizes the operation functions for obtaining the combination of optimal entities which realize the Internet of Things request when a relevant task or operation request is in the presence in the future. Therefore, an equipment interoperation problem under different protocol standard systems can be solved.

Owner:CHINA TELECOM CORP LTD

Named entity identification method based on neural network, computing device and storage medium

PendingCN111160033AAccurate entity recognition resultsImprove accuracySemantic analysisNeural architecturesPattern recognitionNamed-entity recognition

The invention discloses a named entity recognition method based on a neural network, and the method comprises the steps: carrying out the word segmentation of a to-be-recognized corpus, so as to obtain a word segmentation result, wherein a to-be-recognized prediction comprises a plurality of characters; obtaining the character attribute, the word attribute and the part-of-speech attribute of eachcharacter in the to-be-recognized corpus based on the word segmentation result; for each character, constructing an attribute vector representing the character based on the character attribute, the word attribute and the part-of-speech attribute of the character; processing the attribute vector of each word by using a neural network model to obtain an identifier of each word, the identifier indicating that the corresponding word is one of a beginning word of the named entity, a middle word of the named entity, an ending word of the named entity, and a non-named entity word; and determining thenamed entity in the to-be-identified corpus according to the obtained character identifier. The invention further discloses corresponding computing equipment and a storage medium.

Named entity identification method based on neural network, computing device and storage medium

View all

Owner:CHEZHI HULIAN BEIJING SCI & TECH CO LTD

Named entity processing method, system and equipment

PendingCN113011187AImprove accuracyCharacter and pattern recognitionNatural language data processingText categorizationEngineering

The invention provides a named entity processing method, system and equipment. The method comprises the steps of obtaining a to-be-recognized text and text descriptions related to entities in the to-be-recognized text in a knowledge base, matching the to-be-recognized text with the text descriptions of the corresponding entities in the knowledge base through a word segmentation method and a word segmentation method, generating an entity processing result according to a matching result, and determining the association degree between each entity in the to-be-recognized text and the knowledge base text description, so a classification system can obtain the emphasis of the text before classifying the to-be-recognized text, and the text classification accuracy can be improved.

Owner:PING AN TECH (SHENZHEN) CO LTD