35 results about "Name disambiguation" patented technology

The invention relates to a name disambiguation method orienting Chinese authors in English literature. The method mainly comprises the following steps of (1) extracting personnel information of authors in bibliographical reference information of the English literature, and building the collaboration relationship, the reference relationship and the like between the authors; (2) comparing e-mail addresses of the authors with the same names; (3) calculating the similarity of the affiliated units and subjects of the authors with the same names; (4) calculating the similarity of the collaboration relationship of the authors with the same names; (5) calculating the similarity of the reference relationship of the authors with the same names; (6) performing the name disambiguation on the basis of the three-similarity clustering calculated in the step (3) to the step (5).

Name disambiguation by using private / global directories and communication contexts. Exemplary embodiments include a name disambiguation method, including identifying a plurality of names of persons in text received in the computer system, each of the plurality of names including a start and an end position, a family name and a given name, for each of the plurality of names of persons in the text, retrieving personal identification information from a global directory as a candidate for a name match, retrieving personal identification information from a private directory as a candidate for the name match, comparing the private directory personal identification information with the global directory personal identification information, eliminating the personal identification information from the private directory, retrieving frequencies of communications associated with the personal identification information from the private directory and displaying related information for each of the plurality of names of persons in the text.

The invention discloses a method for disambiguating entity names in a medical disease diagnosis record. Based on a heterogeneous concomitant disease network and a graph model, the entity names in the medical disease diagnosis record are disambiguated. The similarity between to-be-disambiguated entity names and candidate entity names is used as local information, and the contribution of other to-be-disambiguated entities in the same record to current to-be-disambiguated entities serves as global information, so that the accuracy of medical entity name disambiguation can be improved; the heterogeneous concomitant disease network is established according to the disease diagnosis record and annotation data, so that the relationships between the diseases and between the disease and the operation can be reflected more intuitively and credibly; and the entity names are subjected to standard name mapping accurately and efficiently, so that the problem of ambiguity of medical disease entity names in diagnosis information is solved, and the practical application demands are met.

A “Name Disambiguator” provides various techniques for implementing an interactive framework for resolving or disambiguating entity names (associated with objects such as publications) for entity searches where two or more same or similar names may refer to different entities. More specifically, the Name Disambiguator uses a combination of user input and automatic models to address the disambiguation problem. In various embodiments, the Name Disambiguator uses a two part process, including: 1) a global SVM trained from large sets of documents or objects in a simulated interactive mode, and 2) further personalization of local SVM models (associated with individual names or groups of names such as, for example, a group of coauthors) derived from the global SVM model. The result of this process is that large sets of documents or objects are rapidly and accurately condensed or clustered into ordered sets by that are organized by entity names.

The invention discloses a scholar name ambiguity elimination method of fusing academic influence. The method comprises: constructing a social network by a disambiguation data sub-set and a source dataset according to coauthoring and citation relationships thereof, and calculating influence of each node in the disambiguation data sub-set according to the network relationships; respectively constructing three network relationships of scholars and scholars, scholars and literature and literature and literature inside the disambiguation data sub-set according to node relationships, and using a sorting-based loss function and combining node influence similarity to jointly learn similarity among scholar nodes in the multiple networks; and constructing a clustering function on the basis of the node similarity and the node influence, and thus realizing a better disambiguation effect. According to the method, the problem of information missing in academic data is overcome while personal privacy is protected, social-network features are fully utilized, the node influence and the node similarity are fused, and the scholar name disambiguation effect is effectively improved.

Name disambiguation by using private / global directories and communication contexts. Exemplary embodiments include a name disambiguation method, including identifying a plurality of names of persons in text received in the computer system, each of the plurality of names including a start and an end position, a family name and a given name, for each of the plurality of names of persons in the text, retrieving personal identification information from a global directory as a candidate for a name match, retrieving personal identification information from a private directory as a candidate for the name match, comparing the private directory personal identification information with the global directory personal identification information, eliminating the personal identification information from the private directory, retrieving frequencies of communications associated with the personal identification information from the private directory and displaying related information for each of the plurality of names of persons in the text.

The invention relates to a paper same-named author disambiguation method based on a high-confidence-degree characteristic attribute hierarchical-clustering method. The paper same-named author disambiguation method mainly comprises the steps that 1, original data is firstly extracted from an academic search engine, characteristic attribute values are extracted, and normalized processing is conducted on the values; 2, another name groups are firstly formed according to rules, and homonymic author ambiguity groups are generated according to the another name groups;3, similarity calculation and disambiguation method selection is performed for characteristic attributes respectively; 4, the high-confidence-degree characteristic attribute hierarchical-clustering method is achieved through attribute confidence-degree assessment performed in the step 3. By applying the paper same-named author disambiguation method, the name disambiguation speed is ensured, and the disambiguation accuracy is also improved.

The invention provides a name disambiguation method and apparatus. The method comprises the following steps: preprocessing full-text information of names to be disambiguated so as to extract semantic features of the full-text information; according to the semantic features, generating semantic fingerprints of the full-text information of the names to be disambiguated, including mail fingerprints, coauthor fingerprints, mechanism fingerprints and text fingerprints; through comparing the full-text information of the names to be disambiguated with semantic fingerprints having same-name full-text information as the names to be disambiguated in a preset semantic fingerprint database, determining similarity between the full-text information of the names to be disambiguated and the semantic fingerprints having the same-name full-text information as the names to be disambiguated in the preset semantic fingerprint database; and according to the semantic fingerprint similarity, determining a name group after disambiguation which the semantic fingerprints of the full-text information of the names to be disambiguated belongs to. By using such a method, while name disambiguation accuracy is ensured, the name disambiguation speed is improved, and increment name disambiguation is supported.

The invention provides a device and a method for name disambiguation clustering. The device for data processing on a name training set comprises the following units: a representative similarity determination unit for determining the representative similarity of the name training set, wherein the representative similarity is a representative value of the inter-textual similarity in the name training set; a preferable similarity threshold selection unit for clustering the name training set by using different similarity thresholds so as to select the similarity threshold which makes the clustering effect better as the preferable similarity threshold; and a function fitting unit for fitting a function which represents the corresponding relation between the representative similarity and the preferable similarity threshold according to the representative similarity and the preferable similarity threshold of each name training set in at least two name training sets.

The present invention discloses a method for detecting the same name of document writers, belonging to the technical field of data mining. The method fully uses a characteristic of same name disambiguation of a single characteristic similarity and single characteristic fusion in scientific literature. The method includes the steps of firstly modeling for a to-be-used document, then, calculating a similarity of every two single characteristics by using a single characteristic similarity detection method, and calculating identification capability of each single characteristic by using a disambiguation method based on the single characteristic similarity, so as to design multi-characteristic fusion disambiguation rules, and provide a method for detecting the same name of the document writers. The detection method integrates advantages of single characteristics of disambiguating the physical writer names, so that the method has high accuracy and callback rate in identification.

The invention discloses a dictionary and semantic disambiguation-based name recognition method and device. The method comprises the steps of name extraction and name disambiguation. The device comprises a name extraction module and a name disambiguation module. According to the name recognition method provided by the invention, all the possible names are found out according to a name dictionary, name disambiguation is carried out through a minimum clearance and a shortest word segmentation length, and finally correct name information is obtained, so that the problems that the names cannot be distinguished and the names are excessively recognized as the word segmentation is incorrect are avoided, the Chinese semantic recognition correctness is improved and the work efficiency of related application personnel is improved.

The invention relates to a disambiguation processing method for a cross-enterprise personnel name duplication phenomenon in industrial and commercial registration information, and the method comprises the steps: carrying out the data collection and filtering processing according to the industrial and commercial registration information, and obtaining an industrial and commercial information personnel list; sampling the obtained business information personnel list to obtain part of personnel information data and corresponding enterprise registration information; grouping the obtained data by constructing an undirected graph model, and calculating the similarity between every two nodes in each sub-graph generated by the undirected graph model; and according to the training vector and the prediction vector, constructing a similarity vector to train a logic regression model, and carrying out similarity weighting processing to obtain a name disambiguation result. The invention also relates to a corresponding system, device, processor and storage medium. By adopting the method, the system, the device, the processor and the storage medium, the enterprise names can be automatically disambiguated, and a certain support is provided for enterprise association relationship analysis.

The invention discloses a name ambiguity eliminating method applied to Web task searching, which comprises the following steps of S1, extracting an HTML webpage source code, and extracting noise irrelevant to character information from the HTML webpage source code; S2, extracting a character webpage feature set; S3, generating a combined feature vector representing a certain person related webpagefrom the person webpage feature set extracted in the step S2; S4, performing hierarchical clustering by adopting a condensation hierarchical clustering algorithm to obtain a character webpage clustering result. According to the method, through introduction of the n-element capital model, the limitation of traditional named entity recognition is solved, named entity extraction is limited, and a plurality of special vocabularies and special vocabularies in the text cannot be extracted; different extracted features are endowed with different weights according to the importance of the features tothe character representation, so that the name disambiguation accuracy is improved.

The invention provides a LightGBM classification and representation learning-based name disambiguation method and a LightGBM classification and representation learning-based name disambiguation system for scientific literature data and aiming at author homonymy phenomena in literatures. According to the supervised learning part, meta-information features of papers in a training set and associated information features among the papers are extracted by utilizing feature engineering, a positive example and negative example sample pair data set is constructed through sampling and serves as input of a LightGBM dichotomy model, and model output serves as the probability that the two papers belong to the same author. The representation learning part refers to a word2vec text semantic representation method and a meta-path-based relation network representation method to capture semantic information of papers and relation characteristics between the papers. And finally, based on the output of the supervision model and the representation learning model, cluster division is performed on the to-be-disambiguated paper set by using a hierarchical clustering algorithm to realize homonymy disambiguation. According to the method, high expandability and stability can be achieved on the premise that the accuracy rate and the recall rate are not lost, parallel calculation can be completely achieved, and the execution efficiency is improved.

The invention relates to the artificial intelligence technology, and discloses a paper author disambiguation method comprising the following steps: respectively forming author names involved in all papers in a database into name trees according to preset rules; obtaining association relationship heterogeneous networks corresponding to all papers in a database; obtaining paper semantic representations respectively corresponding to all papers in the database; constructing a similar matrix based on the name tree, the association relationship heterogeneous network and the paper semantic representation; clustering the similar matrixes to obtain paper clustering groups corresponding to all papers in a database; judging whether the paper clustering group corresponding to the author to be disambiguated belongs to a paper clustering group corresponding to a specified author or or not; and if not, judging that the author to be disambiguated is different from the specified author. According to the method and device, the author names are preprocessed to construct the name tree, then clustering errors caused by different expression modes of name writing are eliminated according to the name tree, it is guaranteed that the names of the same author are divided into the same group as much as possible, and the name disambiguation accuracy is improved.

The invention discloses a part-of-speech tagging-based internet news related place name identification method and a part-of-speech tagging-based internet news related place name identification system,and belongs to the technical field of natural language processing. According to the part-of-speech tagging-based internet news related place name identification method, the method comprises the stepsof supplementing the context information of news by utilizing the overall reporting region of a news media column; assisting a place name disambiguation program to correctly judge a place name, converting news contents into a pure noun phrase sequence by utilizing part-of-speech tagging, carrying out place name identification on the noun phrase sequence, carrying out place name subtraction on place name identification results twice, eliminating inaccurate place names, and finally carrying out weighted summary on the two place name subtraction results to confirm the place name. The part-of-speech tagging-based internet news related place name identification method is popular and easy to understand. The implementation process of the method is simple. The problem that news related place names are low in extraction accuracy can be effectively solved. The part-of-speech tagging-based internet news related place name identification method has good application and popularization value.

The invention discloses a method and a device for author naming disambiguation and electronic equipment. The method comprises the steps of determining a unique author of a paper from an academic dataset by utilizing a pre-trained classification model according to related information of the paper; searching the academic data set to obtain an alternative paper set by utilizing related information of papers for papers for which a unique author cannot be determined; and clustering the papers in the alternative paper set to obtain a plurality of categories, performing reverse classification on thepapers in the alternative paper set to determine the categories where the papers are located, and creating unique authors for the papers according to the categories. In actual work, the method provided by the invention is adopted to perform naming disambiguation on a big data set, and an efficient and extensible effect is achieved on the premise of not losing recall and precision. Therefore, themethod provided by the invention provides an effective solution for naming disambiguation of an oversized data set.

The invention provides a device and a method for name disambiguation clustering. The device for data processing on a name training set comprises the following units: a representative similarity determination unit for determining the representative similarity of the name training set, wherein the representative similarity is a representative value of the inter-textual similarity in the name training set; a preferable similarity threshold selection unit for clustering the name training set by using different similarity thresholds so as to select the similarity threshold which makes the clustering effect better as the preferable similarity threshold; and a function fitting unit for fitting a function which represents the corresponding relation between the representative similarity and the preferable similarity threshold according to the representative similarity and the preferable similarity threshold of each name training set in at least two name training sets.

The embodiment of the invention discloses a named disambiguation method and device and a computer readable storage medium. The named disambiguation accuracy can be improved. The method comprises the steps of extracting a single piece of information from an external information source, extracting a keyword from the single piece of information, inquiring in a local library through the keyword to obtain M results with the highest matching degree, and performing naming and disambiguation on the M results with the highest matching degree in the local library according to the single piece of information. According to the method, under the condition that an external information source is introduced, the external information source is used as an important support of a local library, and the locallibrary is combined with the external information source, so that the problem of named entity indexing errors existing in the local library is solved, and the named disambiguation accuracy is improved.

The invention discloses an enterprise name disambiguation method and device, electronic equipment and a storage medium, and relates to the technical field of knowledge maps. According to the specificimplementation scheme, the method comprises the steps of obtaining an enterprise name abstract in a pre-constructed enterprise name abstract set, wherein the enterprise name abstract set comprises enterprise name abstracts corresponding to all enterprises; searching each enterprise name abstract in the news; if at least one enterprise name abstract is found in the news, obtaining an enterprise name corresponding to each found enterprise name abstract in the news; and if at least one enterprise name abbreviation or alias exists in the enterprise names obtained from the news, disambiguating theenterprise name abbreviation or alias obtained from the news according to the text features of the news and the pre-obtained enterprise information of each enterprise. According to the embodiment of the invention, the enterprise name abbreviation or alias appearing in the news can be efficiently disambiguated, so that the news and related enterprises can be quickly aggregated.

The present invention discloses a natural person name disambiguation method, device and medium based on enterprise association relationship, the method includes: obtaining basic training data, using enterprise and personnel as two node types, and constructing a basic heterogeneous graph of enterprise-personnel ;According to the preset splitting rules, split some personnel nodes with multiple edges in the basic heterogeneous graph to obtain a derivative graph; according to the derivative graph, train the preset heterogeneous graph neural network model , to obtain a node vector representation model; add the personnel to be merged into the basic heterogeneous graph as a personnel node, and according to the node vector representation model, judge that the personnel node to be merged and other nodes with the same name in the basic heterogeneous graph Whether to merge. Compared with the prior art, the method of the present invention graphs the enterprise data through the enterprise association relationship, and then disambiguates the nodes with the same name through the trained graph neural network model, which greatly improves the accuracy of the disambiguation results.

The invention provides a name disambiguation method and apparatus. The method comprises the following steps: preprocessing full-text information of names to be disambiguated so as to extract semantic features of the full-text information; according to the semantic features, generating semantic fingerprints of the full-text information of the names to be disambiguated, including mail fingerprints, coauthor fingerprints, mechanism fingerprints and text fingerprints; through comparing the full-text information of the names to be disambiguated with semantic fingerprints having same-name full-text information as the names to be disambiguated in a preset semantic fingerprint database, determining similarity between the full-text information of the names to be disambiguated and the semantic fingerprints having the same-name full-text information as the names to be disambiguated in the preset semantic fingerprint database; and according to the semantic fingerprint similarity, determining a name group after disambiguation which the semantic fingerprints of the full-text information of the names to be disambiguated belongs to. By using such a method, while name disambiguation accuracy is ensured, the name disambiguation speed is improved, and increment name disambiguation is supported.

The invention belongs to the technical field of name disambiguation, and particularly relates to a Chinese and English literature author name disambiguation method. According to the method, Chinese author name disambiguation and English author name disambiguation are carried out based on semantic fingerprints, author cooperation network similarity, author reference network similarity and the like, and disambiguation of Chinese authors and name pinyin in English literatures is completed according to a Chinese disambiguation result and an English disambiguation result. According to the method, whether authors of different literatures are the same person or not can be accurately distinguished, the same author in Chinese and English can be well recognized, the author needing to be found can be quickly positioned, the accuracy rate is high, and retrieval work can be conveniently carried out; the calculation of the similarity of the scientific research duration of the authors is introduced, so that disambiguation of Chinese and English names of the Chinese authors can be well assisted, the age range of the authors can be determined, other authors with the same name not in the range can be filtered out, and the disambiguation accuracy is improved.

The invention discloses a natural person name disambiguation method and device based on an enterprise association relationship and a medium, and the method comprises the steps: obtaining basic training data, taking an enterprise and a person as two node types, and constructing an enterprise-person basic heterogeneous graph; according to a preset splitting rule, splitting partial personnel nodes with a plurality of edges in the basic heterogeneous graph to obtain a derivative graph; according to the derivative graph, training a preset heterogeneous graph neural network model to obtain a node vector representation model; and adding to-be-merged personnel as personnel nodes into the basic heterogeneous graph, and judging whether the to-be-merged personnel nodes and other homonymous nodes in the basic heterogeneous graph need to be merged or not according to the node vector representation model. Compared with the prior art, the method has the advantages that the enterprise data is graphical through the enterprise association relationship, and then the ambiguity elimination processing is performed on the homonymous nodes through the trained graph neural network model, so that the accuracy of the ambiguity elimination result is greatly improved.

The invention discloses a travel note place name disambiguation method based on time geography. The method comprises the following steps: 1) extracting place names and time labels thereof in a travelnote text, dividing the extracted place names into ambiguous place names and unambiguous place names, allocating unique longitude and latitude positions to the unambiguous place names, and listing allpossible longitude and latitude positions corresponding to the ambiguous place names; 2) disambiguating the ambiguous place names by using PPA; 3) disambiguating the ambiguous place names by utilizing the reachable domain at the determined moment; and 4) sorting by using probability time geography, calculating a probability for each remaining ambiguous place name, and performing descending sortaccording to a calculation result. The invention provides a disambiguation method based on time geography, which is different from previous methods based on rules and the like, is suitable for travelnote place name disambiguation, supplements the disambiguation method in the aspect of fine-grained place names, and enables place name disambiguation to be more accurate.

The embodiment of the present application relates to the field of information retrieval technology, and discloses a name disambiguation method, device, electronic equipment, and computer-readable storage medium, wherein the name disambiguation method includes: according to the word sparse distributed representation generated in advance based on the training corpus SDR, determine the document information of at least two documents in at least two language categories to be disambiguated, one document corresponds to one language category; then, based on the pre-built document author classification model for at least two language categories, According to the document information of each document in at least two language types, classify each document according to the author of the document to obtain the first author category corresponding to each document, and the document author classification model of one language type corresponds to the processing Documents of corresponding language categories; Next, the first author categories under each language category are merged, so as to disambiguate the names of the document authors of each document in each language category.

The invention relates to a name disambiguation method orienting Chinese authors in English literature. The method mainly comprises the following steps of (1) extracting personnel information of authors in bibliographical reference information of the English literature, and building the collaboration relationship, the reference relationship and the like between the authors; (2) comparing e-mail addresses of the authors with the same names; (3) calculating the similarity of the affiliated units and subjects of the authors with the same names; (4) calculating the similarity of the collaboration relationship of the authors with the same names; (5) calculating the similarity of the reference relationship of the authors with the same names; (6) performing the name disambiguation on the basis of the three-similarity clustering calculated in the step (3) to the step (5).