Entity Disambiguation Research Articles

Progressive Joint Framework for Chinese Question Entity Discovery and Linking With Question Representations

Chinese question entity discovery and linking (QEDL) may encounter short texts and small-scale annotated datasets, which may invalidate certain machine learning algorithms. In this paper, we propose a progressive joint framework for Chinese QEDL, which leverages the mutual dependency information of these two tasks to enhance the performance with each other. The framework uses the candidate entity generation (CEG) of entity linking to iteratively augment the overall process of entity discovery that consists of mention generation, filtering and merging modules. In mention generation module, to reduce the hand-crafted effort of the rule-based entity discovery, we develop a question representation method to generate domain-independent entity discovery rules, and use CEG to check the extracted mentions in priority order. This module can embed extracted mentions into other entity discovery methods as one feature or as extra mentions to alleviate insufficiencies of annotated datasets. The mentions filtering module leverages the joint features of extracted mentions and CEG’s entities to build a voting model and filter out low-confidence mentions. Moreover, the mentions merging module merges different patterns’ mention-entity pairs and check their corresponding candidate entities with CEG. During entity linking, we incorporate the joint features of questions, extracted mentions and CEG’s entities into a ranking model for entity disambiguation. Finally, we conduct experiments on two real datasets and compare our approach with other state-of-the-art methods. The results illustrate that the proposed framework can reduce error accumulation and flexibly combine different entity discovery methods, which significantly improves the performance on small-scale datasets.

Named Entity Disambiguation Anaphora Resolution and Question Answering in Speech

Named Entity Disambiguation Anaphora Resolution and Question Answering in Speech

Query expansion using named entity disambiguation for a question‐answering system

SummaryDue to the exponential growth in the size of available data, people often prefer direct answers in many cases, instead of a set of ranked documents while searching using a search engine. Furthermore, people often do not provide enough information regarding their search. Query expansion would help in such cases so that queries would be best understood by search engines. Although the queries are expanded, many a times, they end up with wrong results due to the entity mentions in the questions. Hence, we need to disambiguate named entities properly so as to choose the right mention. In this paper, a method is proposed for named entity disambiguation in order to help in query expansion. An adapted Lesk similarity measure is used to find the similarity between questions and the target Wikipedia disambiguation pages. The proposed approach has shown very good results in disambiguating organization‐ and miscellaneous‐type entity mentions.

Named Entity Disambiguation Based on Classified and Structural Semantic Relatedness

Named Entity Disambiguation Based on Classified and Structural Semantic Relatedness

Named Entity Extraction for Knowledgebase Enhancement

Past decade witnessed an explosive growth in the amount of unstructured data, especially in the public domain, mainly due to Web 2.0 and social media. This led to the creation of applications, called information extractors, that extract structured information from unstruc- tured data. The extracted information is stored in a Knowledge Base (KB). KB stores facts about entities like name, type and other attributes. My PhD thesis entitled 'Named Entity Extraction for Knowledgebase Enhancement' deals with information extraction on named entities with the purpose of enhancing a KB. The enhanced KB is in turn used by the information extraction task to refine the extraction process. Thus, KB provides structure and guidance to the extraction task, and gets enhanced by the results of the extraction task. Here we see that the tasks of entity extraction and KB enhancement are mutually dependent and mutually beneficial. Hence in my research I propose methods to enhance both the tasks, in an effort to build a strong and sound named entity extraction system. Named Entity Extraction, also known as Entity Linking (EL) in scientific literature, is the task of determining the identity of entities mentioned in text. EL helps automatic extraction of structured information about entities from unstructured data, which is stored in the KB. EL consists of Mention Detection and Entity Disambiguation. In my research, I propose methods to enhance mention detection, entity disambiguation and KB enhancement.

Report on the 2nd International Workshop on Recent Trends in News Information Retrieval (NewsIR'18)

The news industry has undergone a revolution in the past decade, with substantial changes continuing to this day. News consumption habits are changing due to the increase in the volume of news and the variety of sources. Readers need new mechanisms to cope with this vast volume of information in order to not only find a signal in the noise, but also to understand what is happening in the world given the multiple points of view describing events. These challenges in journalism relate to Information Retrieval (IR) and Natural Language Processing (NLP) fields such as: verification of a source's reliability; the integration of news with other sources of information; real-time processing of both news content and social streams; de-duplication of stories; and entity detection and disambiguation. Although IR and NLP have been applied to news for decades, the changing nature of the space requires fresh approaches and a closer collaboration with our colleagues from the journalism environment. Following the success of the previous version of the workshop (NewsIR'16), the goal of this workshop, held in conjunction with ECIR 2018, is to continue to stimulate such discussion between the communities and to share interesting approaches to solve real user problems. A total number of 19 submissions were received and reviewed, of which 12 were accepted for presentation. In addition to that, we had over 30 registered participants in the workshop who were pleased to attend the two keynote talks given by well-known experts in the field - Edgar Meij (from industry) and Peter Tolmie (from academia) and oral and poster presentations from the accepted papers. The workshop also included a breakout session to discuss ideas for a future data challenge in news IR and closed with a focused panel discussion to reflect on the day. In summary, several ideas were presented in the workshop on solving complex information needs in the news domain. In addition, the workshop concluded with suggestions of important challenges and shared tasks to work on as a community for News IR.

BTM Topic Modeling Approach to Named Entity Linking

Named entity link is the process of linking a given named entity in a document to an unambiguous entity in a knowledge base, including the merging of synonymous entities, disambiguation of ambiguous entities, etc. Firstly, the paper construct the named entity knowledge base by using the offline version of Wikipedia. We improve the MPM co word measure which only considers the co-occurrence frequency without considering the frequency of individual words so that it can calculate word co-occurrence coefficient. Secondly, we propose a method named entity BTM link based on topic model. Finally, the named entity in the given text is linked to an ambiguous named entity in the knowledge base.

Robust named entity disambiguation with random walks

Robust named entity disambiguation with random walks

Entity linking of tweets based on dominant entity candidates

Entity linking, also known as semantic annotation, of textual content has received increasing attention. Recent works in this area have focused on entity linking on text with special characteristics such as search queries and tweets. The semantic annotation of tweets is specially proven to be challenging given the informal nature of the writing and the short length of the text. In this paper, we propose a method to perform entity linking on tweets built based on one primary hypothesis. We hypothesize that while there are formally many possible entity candidates for an ambiguous mention in a tweet, as listed on the disambiguation page of the corresponding entity on Wikipedia, there are only few entity candidates that are likely to be employed in the context of Twitter. Based on this hypothesis, we propose a method to identify such dominant entity candidates for each ambiguous mention and use them in the annotation process. Particularly, our proposed work integrates two phases (i) dominant entity candidate detection, which applies community detection methods for finding the dominant candidates of ambiguous mentions; and (ii) named entity disambiguation that links a tweet to entities in Wikipedia by only considering the identified dominant entity candidates. Our investigations show that: (1) there are only very few entity candidates for each ambiguous mention in a tweet that need to be considered when performing disambiguation. This helps us limit the candidate search space and hence noticeably reduce the entity linking time; (2) limiting the search space to only a subset of disambiguation options will not only improve entity linking execution time but will also lead to improved accuracy of the entity linking process when the main entity candidates of each mention are mined from a temporally aligned corpus. We show that our proposed method offers competitive results with the state-of-the-art methods in terms of precision and recall on widely used gold standard datasets while significantly reducing the time for processing each tweet.

Location Extraction from Social Media

Location extraction, also called “toponym extraction,” is a field covering geoparsing, extracting spatial representations from location mentions in text, and geotagging, assigning spatial coordinates to content items. This article evaluates five “best-of-class” location extraction algorithms. We develop a geoparsing algorithm using an OpenStreetMap database, and a geotagging algorithm using a language model constructed from social media tags and multiple gazetteers. Third-party work evaluated includes a DBpedia-based entity recognition and disambiguation approach, a named entity recognition and Geonames gazetteer approach, and a Google Geocoder API approach. We perform two quantitative benchmark evaluations, one geoparsing tweets and one geotagging Flickr posts, to compare all approaches. We also perform a qualitative evaluation recalling top N location mentions from tweets during major news events. The OpenStreetMap approach was best (F1 0.90+) for geoparsing English, and the language model approach was best (F1 0.66) for Turkish. The language model was best (F1@1km 0.49) for the geotagging evaluation. The map database was best (R@20 0.60+) in the qualitative evaluation. We report on strengths, weaknesses, and a detailed failure analysis for the approaches and suggest concrete areas for further research.

Task-Specific Representation Learning for Web-Scale Entity Disambiguation

Named entity disambiguation (NED) is a central problem in information extraction. The goal is to link entities in a knowledge graph (KG) to their mention spans in unstructured text. Each distinct mention span (like John Smith, Jordan or Apache) represents a multi-class classification task. NED can therefore be modeled as a multitask problem with tens of millions of tasks for realistic KGs. We initiate an investigation into neural representations, network architectures, and training protocols for multitask NED. Specifically, we propose a task-sensitive representation learning framework that learns mention dependent representations, followed by a common classifier. Parameter learning in our framework can be decomposed into solving multiple smaller problems involving overlapping groups of tasks. We prove bounds for excess risk, which provide additional insight into the problem of multi-task representation learning. While remaining practical in terms of training memory and time requirements, our approach outperforms recent strong baselines, on four benchmark data sets.

Mention and Entity Description Co-Attention for Entity Disambiguation

For the task of entity disambiguation, mention contexts and entity descriptions both contain various kinds of information content while only a subset of them are helpful for disambiguation. In this paper, we propose a type-aware co-attention model for entity disambiguation, which tries to identify the most discriminative words from mention contexts and most relevant sentences from corresponding entity descriptions simultaneously. To bridge the semantic gap between mention contexts and entity descriptions, we further incorporate entity type information to enhance the co-attention mechanism. Our evaluation shows that the proposed model outperforms the state-of-the-arts on three public datasets. Further analysis also confirms that both the co-attention mechanism and the type-aware mechanism are effective.

DISC: Disambiguating homonyms using graph structural clustering

Author name ambiguity degrades information retrieval, database integration, search results and, more importantly, correct attributions in bibliographic databases. Some unresolved issues include how to ascertain the actual number of authors, how to improve the performance and how to make the method more effective in terms of representative clustering metrics (average cluster purity, average author purity, K-metric, pairwise precision, pairwise recall, pairwise-F1, cluster precision, cluster recall and cluster-F1). It is a non-trivial task to disambiguate authors using only the implicit bibliographic information. An effective method ‘DISC’ is proposed that uses graph community detection algorithm, feature vectors and graph operations to disambiguate homonyms. The citation data set is pre-processed and ambiguous author blocks are formed. A co-authors graph is constructed using authors and their co-author’s relationships. A graph structural clustering ‘gSkeletonClu’ is applied to identify hubs, outliers and clusters of nodes in a co-author’s graph. Homonyms are resolved by splitting these clusters of nodes across the hub if their feature vector similarity is less than a predefined threshold. DISC utilises only co-authors and titles that are available in almost all bibliographic databases. With little modifications, DISC can also be used for entity disambiguation. To validate the DISC performance, experiments are performed on two Arnetminer data sets and compared with five previous unsupervised methods. Despite using limited bibliographic metadata, DISC achieves on average K-metric, pairwise-F1, and cluster-F1 of 92%, 84% and 74%, respectively, using Arnetminer-S and 86%, 80% and 57%, respectively, using Arnetminer-L. About 77.5% and 73.2% clusters are within the range (ground truth clusters ± 3) in Arnetminer-S and Arnetminer-L, respectively.

Exploiting semantic similarity for named entity disambiguation in knowledge graphs

With the increasing popularity of large scale Knowledge Graph (KG)s, many applications such as semantic analysis, search and question answering need to link entity mentions in texts to entities in KGs. Because of the polysemy problem in natural language, entity disambiguation is thus a key problem in current research. Existing disambiguation methods have considered entity prominence, context similarity and entity-entity relatedness to discriminate ambiguous entities, which are mainly working on document or paragraph level texts containing rich contextual information, and based on lexical matching for computing context similarity. When meeting short texts containing limited contextual information, such as web queries, questions and tweets, those conventional disambiguation methods are not good at handling single entity mention and measuring context similarity. In order to enhance the performance of disambiguation methods based on context similarity with such short texts, we propose SCSNED method for disambiguation based on semantic similarity between contextual words and informative words of entities in KGs. Specially, we exploit the effectiveness of both knowledge-based and corpus-based semantic similarity methods for entity disambiguation with SCSNED. Moreover, we propose a Category2Vec embedding model based on joint learning of word and category embedding, in order to compute word-category similarity for entity disambiguation. We show the effectiveness of these proposed methods with illustrative examples, and evaluate their effectiveness in a comparative experiment for entity disambiguation in real world web queries, questions and tweets. The experimental results have identified the effectiveness of different semantic similarity methods, and demonstrated the improvement of semantic similarity methods in SCSNED and Category2Vec over the conventional context similarity baseline. We further compare the proposed approaches with the state of the art entity disambiguation systems and show the performances of the proposed approaches are among the best performing systems. In addition, one important feature of the proposed approaches using semantic similarity, is the potential application on any existing KGs since they mainly use common features of entity descriptions and categories. Another contribution of the paper is an updated survey on background of entity disambiguation in KGs and semantic similarity methods.

Entity disambiguation with memory network

Abstract We develop a computational approach based on memory network for entity disambiguation. The approach automatically finds important clues of a mention from surrounding contexts with attention mechanism, and leverages these clues to facilitate entity disambiguation. Unlike existing feature-based methods, this approach does not rely on any manually designed features. Unlike existing neural models such as recurrent or convolutional neural network, this approach leverages the importance of context words in an explicit way. The model could be easily trained with back-propagation. To effectively learn the model parameters, we automatically collect large-scale mention-entity pairs from Wikipedia as training data. We verify the effectiveness of the proposed approach on a benchmark dataset from TAC-KBP evaluation in 2010. Experimental results demonstrate that our approach empirically surpasses strong feature based and neural network based methods. Model analysis further reveals that our approach has the capacity to discover important clues from contexts.

EREL: an Entity Recognition and Linking algorithm

ABSTRACTThis paper introduces the EREL algorithm that integrates Entity Recognition, Co-reference Resolution (CR) and Disambiguation. The algorithm recognizes entity mentions as the longest name based on the name dictionary constructed from the Wikipedia data. The CR is integrated into the algorithm to improve the performance in processing short-form or abbreviated names. The algorithm employs a new approach in disambiguation entities using new features as entity-level context information and case-sensitive data about the mention in disambiguation. Tested on four benchmark data sets in the GERBIL framework, EREL outperforms the current Entity Linking methods. EREL achieves the micro f-score as 0.83 in both tasks Disambiguate to Wikipedia and Annotate to Wikipedia.

Entity disambiguation with decomposable neural networks

Entity disambiguation is a fundamental task in natural language processing and computational linguistics. Given a query consisting of a mention (name string) and a background document, entity disambiguation aims at linking the mention to an entity from a reference knowledge base such as Wikipedia. A main challenge of this task is how to effectively represent the meaning of the mention and the entity, based on which the semantic relatedness between the mention and the entity could be conveniently measured. Towards this goal, we introduce computational models to effectively represent the mention and the entity in some vector space. We decompose the problem into subproblems and develop various neural network architectures, all of which are purely data‐driven and capable of learning continuous representations of the mention and the entity from data. To effectively train the neural network models, we explore a simple yet effective way that enables us to collect millions of training examples from Wikipedia without using any manual annotation. Empirical results on two benchmark datasets show that our approaches based on convolutional neural network and long short‐term memory consistently outperform top‐performed systems on both datasets. WIREs Data Mining Knowl Discov 2017, 7:e1215. doi: 10.1002/widm.1215This article is categorized under: Algorithmic Development > Text Mining Algorithmic Development > Web Mining

Named entity disambiguation for questions in community question answering

Named entity disambiguation (NED) refers to the task of mapping entity mentions in running texts to the correct entries in a specific knowledge base (e.g., Wikipedia). Although there has been a lot of work on NED for long and formal texts like Wikipedia and news, the task is not well studied for questions in community question answering (CQA). The challenges of the task include little context for mentions in questions, lack of ground truth for learning, and language gaps between CQA and knowledge bases. To overcome these problems, we propose a topic modelling approach to NED for questions. Our model performs learning in an unsupervised manner, but can take advantage of weak supervision signals estimated from the metadata of CQA and knowledge bases. The signals can enrich the context of mentions in questions, and bridge the language gaps between CQA and knowledge bases. Besides these advantages, our model simulates people’s behavior in CQA and thus is intuitively interpretable. We conduct experiments on both Chinese and English CQA data. The experimental results show that our method can significantly outperform state-of-the-art methods when we apply them to questions in CQA.

An easy-to-use evaluation framework for benchmarking entity recognition and disambiguation systems

Entity recognition and disambiguation (ERD) is a crucial technique for knowledge base population and information extraction. In recent years, numerous papers have been published on this subject, and various ERD systems have been developed. However, there are still some confusions over the ERD field for a fair and complete comparison of these systems. Therefore, it is of emerging interest to develop a unified evaluation framework. In this paper, we present an easy-to-use evaluation framework (EUEF), which aims at facilitating the evaluation process and giving a fair comparison of ERD systems. EUEF is well designed and released to the public as an open source, and thus could be easily extended with novel ERD systems, datasets, and evaluation metrics. It is easy to discover the advantages and disadvantages of a specific ERD system and its components based on EUEF. We perform a comparison of several popular and publicly available ERD systems by using EUEF, and draw some interesting conclusions after a detailed analysis.

Entity Disambiguation with Markov Logic Network Knowledge Graphs

Entity Disambiguation with Markov Logic Network Knowledge Graphs