Filters








38,316 Hits in 4.8 sec

Named Entity Recognition with Bidirectional LSTM-CNNs [article]

Jason P.C. Chiu, Eric Nichols
2016 arXiv   pre-print
employ heavy feature engineering, proprietary lexicons, and rich entity linking information.  ...  By using two lexicons constructed from publicly-available sources, we establish new state of the art performance with an F1 score of 91.62 on CoNLL-2003 and 86.28 on OntoNotes, surpassing systems that  ...  The authors would like to thank Collobert et al. (2011b) for releasing SENNA with its word vectors and lexicon, the torch7 framework contributors, and Andrey Karpathy for the reference LSTM implementation  ... 
arXiv:1511.08308v5 fatcat:res6fbe7zvg6llf2axriltiq6y

ETM: Entity Topic Models for Mining Documents Associated with Entities

Hyungsul Kim, Yizhou Sun, Julia Hockenmaier, Jiawei Han
2012 2012 IEEE 12th International Conference on Data Mining  
In this paper, we therefore introduce a novel Entity Topic Model (ETM) for documents that are associated with a set of entities.  ...  However, documents are also associated with further information, such as the set of real-world entities mentioned in them.  ...  Based on P (e|z, E, Z, Φ) and P (z|e, E, Z, Φ), we can rank entities for each topic, and rank topics for each entity. Table VI shows two topics and their entity rankings.  ... 
doi:10.1109/icdm.2012.107 dblp:conf/icdm/KimSHH12 fatcat:wve5vdi4s5cghf53frkbpcwv4e

Named Entity Translation with Web Mining and Transliteration

Long Jiang, Ming Zhou, Lee-Feng Chien, Cheng Niu
2007 International Joint Conference on Artificial Intelligence  
This paper presents a novel approach to improve the named entity translation by combining a transliteration approach with web mining, using web information as a source to complement transliteration, and  ...  A Maximum Entropy model is employed to rank translation candidates by combining pronunciation similarity and bilingual contextual co-occurrence.  ...  WM-NE: Candidate Collection by Searching the Web with NE Same with [Wang et al., 2004] , we use the input English NE as a query to search for the Chinese pages and extract candidates in the returned top  ... 
dblp:conf/ijcai/JiangZCN07 fatcat:hvcets3mdbecza5vjyq3tyw3sq

Web Searching with Entity Mining at Query Time [chapter]

Pavlos Fafalios, Ioannis Kitsos, Yannis Marketakis, Claudio Baldassarre, Michail Salampasis, Yannis Tzitzikas
2012 Lecture Notes in Computer Science  
In this paper we present a method to enrich the classical web searching with entity mining that is performed at query time.  ...  The results of entity mining (entities grouped in categories) can complement the query answers with useful for the user information which can be further exploited in a faceted search-like interaction scheme  ...  Introduction Entity search engines aim at providing the user with entities and relationships between these entities, instead of providing the user with links to web pages.  ... 
doi:10.1007/978-3-642-31274-8_6 fatcat:odmjudorcvhwbazcq5mpfrbn5u

QuTE: Answering Quantity Queries from Web Tables

Vinh Thinh Ho, Koninika Pal, Gerhard Weikum
2021 Proceedings of the 2021 International Conference on Management of Data  
It comprises methods for automatically extracting entity-quantity facts from web tables, as well as methods for online query processing, with new techniques for query matching and answer ranking.  ...  Quantities are financial, technological, physical and other measures that denote relevant properties of entities, such as revenue of companies, energy efficiency of cars or distance and brightness of stars  ...  It has two major components: 1) extracting entity-quantity facts from tables using machine-learning techniques, and 2) matching queries against an indexed repository of such facts and computing ranked  ... 
doi:10.1145/3448016.3452763 fatcat:yezokhuhkffbthapyhlwy7yzom

Named Entity Recognition with Bidirectional LSTM-CNNs

Jason P.C. Chiu, Eric Nichols
2016 Transactions of the Association for Computational Linguistics  
employ heavy feature engineering, proprietary lexicons, and rich entity linking information.  ...  By using two lexicons constructed from publicly-available sources, we establish new state of the art performance with an F1 score of 91.62 on CoNLL-2003 and 86.28 on OntoNotes, surpassing systems that  ...  The authors would like to thank Collobert et al. (2011b) for releasing SENNA with its word vectors and lexicon, the torch7 framework contributors, and Andrey Karpathy for the reference LSTM implementation  ... 
doi:10.1162/tacl_a_00104 fatcat:j7xkgc4n3jdztoyackcgbxy2ne

Enhancing Knowledge Bases with Quantity Facts

Vinh Thinh Ho, Daria Stepanova, Dragan Milchevski, Jannik Strötgen, Gerhard Weikum
2022 Proceedings of the ACM Web Conference 2022  
Prior work on extracting quantity facts from web contents focused on high precision for top-ranked outputs, but did not tackle the KB coverage issue.  ...  , and more.  ...  Extrinsic Use Case: Quantity Search We aim to compute top-ranked answers with quantity filter conditions such as: "buildings with height above 1000 ft", or "sprinters who ran 100 meters under 9.9 seconds  ... 
doi:10.1145/3485447.3511932 fatcat:hkyzkbg32zb6po7467kbfw5a7y

Extracting Contextualized Quantity Facts from Web Tables

Vinh Thinh Ho, Koninika Pal, Simon Razniewski, Klaus Berberich, Gerhard Weikum
2021 Proceedings of the Web Conference 2021  
Quantity queries, with filter conditions on quantitative measures of entities, are beyond the functionality of search engines and QA assistants.  ...  This involves recognizing quantities, with normalized values and units, aligning them with the proper entities, and contextualizing these pairs with informative cues to match sophisticated queries with  ...  Web Tables Entity Answers Quantity Query Text Corpus Qfact Scoring Joint CA & EL Entity Column Linking Alignment Contextualized Qfacts Qfact Matching Qfact Corrob-oration Qfact Extraction Search & Ranking  ... 
doi:10.1145/3442381.3450072 fatcat:p7l4as6amre33kqdlkua7kwj64

Reducing Quantity Hallucinations in Abstractive Summarization [article]

Zheng Zhao, Shay B. Cohen, Bonnie Webber
2020 arXiv   pre-print
The system learns to recognize and verify quantity entities (dates, numbers, sums of money, etc.) in a beam-worth of abstractive summaries produced by state-of-the-art models, in order to up-rank those  ...  summaries whose quantity terms are supported by the original text.  ...  Cardenas, and Shashi Narayan for their helpful feedback. We also would like to thank Ronald A.  ... 
arXiv:2009.13312v1 fatcat:c5r2czfquncddak7bmnbna5pxy

QFinder

Satya Almasian, Milena Bruseva, Michael Gertz
2022 Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval  
In this work, we demonstrate QFinder, our quantity-centric framework for ranking search results for queries with quantity constraints.  ...  Quantities shape our understanding of measures and values, and they are an important means to communicate the properties of objects.  ...  QuTE [6, 7] automatically extracts entity-quantity facts from web tables and ranks them.  ... 
doi:10.1145/3477495.3531672 fatcat:c545rdyh3zb37nq3pcguzkk5re

Distributed Entity Disambiguation with Per-Mention Learning [article]

Tiep Mai, Bichen Shi, Patrick K. Nicholson, Deepak Ajwani, Alessandra Sala
2016 arXiv   pre-print
Existing techniques based on global ranking models fail to capture the individual peculiarities of the words and hence, either struggle to meet the accuracy requirements of many real-world applications  ...  To train and validate the hundreds of thousands of learning models for this purpose, we use a Wikipedia hyperlink dataset with more than 170 million labelled annotations.  ...  Thereafter, a densest subgraph is extracted and the senses with maximum scores are selected. Ganea et al.  ... 
arXiv:1604.05875v1 fatcat:rvzhm4ghm5aarbpqvlvovbus5q

SliceNDice: Mining Suspicious Multi-attribute Entity Groups with Multi-view Graphs [article]

Hamed Nilforoshan, Neil Shah
2020 arXiv   pre-print
Finally, we propose the SliceNDice algorithm which enables efficient extraction of highly suspicious entity groups, and demonstrate its practicality in production, in terms of strong detection performance  ...  one another across multiple attributes (sybil accounts created at the same time and location, propaganda spreaders broadcasting articles with the same rhetoric and with similar reshares, etc.)  ...  Finally, we propose the SLICENDICE algorithm which enables efficient extraction of highly suspicious entity groups, and demonstrate its practicality in production, in terms of strong detection performance  ... 
arXiv:1908.07087v3 fatcat:ilwezdakdrhg5akxzpsi4anm3m

Learning to rank for quantity consensus queries

Somnath Banerjee, Soumen Chakrabarti, Ganesh Ramakrishnan
2009 Proceedings of the 32nd international ACM SIGIR conference on Research and development in information retrieval - SIGIR '09  
and entity ranking inadequate Clue from score-vs.  ...  0.382 0.367 0.350 0.330 0.316 RankSVM prec 0.330 0.312 0.298 0.294 0.284 Proximity features as used in entity ranking Proximity between query token and candidate quantity = reciprocal of number of tokens  ... 
doi:10.1145/1571941.1571985 dblp:conf/sigir/BanerjeeCR09 fatcat:tbgupw5wzncldfpapy34ufkbum

Improving Entity Recommendation with Search Log and Multi-Task Learning

Jizhou Huang, Wei Zhang, Yaming Sun, Haifeng Wang, Ting Liu
2018 Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence  
Entity recommendation, providing search users with an improved experience by assisting them in finding related entities for a given query, has become an indispensable feature of today's Web search engine  ...  ranking.  ...  Second, we use the same method and search sessions to extract training data for learning representations of queries and entities for entity recommendation task.  ... 
doi:10.24963/ijcai.2018/571 dblp:conf/ijcai/HuangZSWL18 fatcat:wp6aafzj5nbvjfdkxwmusf2qhu

Novel Techniques for Text Annotation with Wikipedia Entities [chapter]

Christos Makris, Michael Angelos Simos
2014 IFIP Advances in Information and Communication Technology  
Acknowledgments This research has been co-financed by the European Union (European Social Fund -ESF) and Greek national funds through the Operational Program "Education and Lifelong Learning" of the National  ...  Spots Extraction We use a simple yet very effective method for preprocessing the input text fragments for the extraction of candidate Wikipedia entities.  ...  Let Pg(s i ) the set of candidate Wikipedia entity annotations of spot s i . Each state is constituted of lists of candidate Wikipedia entity annotation of each spot sorted by a ranking criterion.  ... 
doi:10.1007/978-3-662-44654-6_50 fatcat:wy2x2vtvrfdd3cvgubfr3tntw4
« Previous Showing results 1 — 15 out of 38,316 results