Browse wiki

Jump to: navigation, search
Ranking related entities: Components and analyses
Abstract Related entity finding is the task of retuRelated entity finding is the task of returning a ranked list of homepages of relevant entities of a specified type that need to engage in a given relationship with a given source entity. We propose a framework for addressing this task and perform a detailed analysis of four core components; co-occurrence models, type filtering, context modeling and homepage finding. Our initial focus is on recall. We analyze the performance of a model that only uses cooccurrence statistics. While it identifies a set of related entities, it fails to rank them effectively. Two types of error emerge: (1) entities of the wrong type pollute the ranking and (2) while somehow associated to the source entity, some retrieved entities do not engage in the right relation with it. To address (1), we add type filtering based on category information available in Wikipedia. To correct for (2), we add contextual information, represented as language models derived from documents in which source and target entities co-occur. To complete the pipeline, we find homepages of top ranked entities by combining a language modeling approach with heuristics based on Wikipedia's external links. Our method achieves very high recall scores on the end-to-end task, providing a solid starting point for expanding our focus to improve precision; additional heuristics lead to state-of-the-art performance.tics lead to state-of-the-art performance.
Abstractsub Related entity finding is the task of retuRelated entity finding is the task of returning a ranked list of homepages of relevant entities of a specified type that need to engage in a given relationship with a given source entity. We propose a framework for addressing this task and perform a detailed analysis of four core components; co-occurrence models, type filtering, context modeling and homepage finding. Our initial focus is on recall. We analyze the performance of a model that only uses cooccurrence statistics. While it identifies a set of related entities, it fails to rank them effectively. Two types of error emerge: (1) entities of the wrong type pollute the ranking and (2) while somehow associated to the source entity, some retrieved entities do not engage in the right relation with it. To address (1), we add type filtering based on category information available in Wikipedia. To correct for (2), we add contextual information, represented as language models derived from documents in which source and target entities co-occur. To complete the pipeline, we find homepages of top ranked entities by combining a language modeling approach with heuristics based on Wikipedia's external links. Our method achieves very high recall scores on the end-to-end task, providing a solid starting point for expanding our focus to improve precision; additional heuristics lead to state-of-the-art performance.tics lead to state-of-the-art performance.
Bibtextype inproceedings  +
Doi 10.1145/1871437.1871574  +
Has author Bron M. + , Balog K. + , Maarten de Rijke +
Has extra keyword Co-occurrence + , Co-occurrence statistics + , Context modeling + , Contextual information + , End-to-end task + , Entity search + , External links + , Four-core + , Home page + , Language model + , Language modeling + , State-of-the-art performance + , Wikipedia + , Knowledge management + , Computational linguistics +
Has keyword Entity search + , Language modeling + , Wikipedia +
Isbn 9781450300995  +
Language English +
Number of citations by publication 0  +
Number of references by publication 0  +
Pages 1079–1088  +
Published in International Conference on Information and Knowledge Management, Proceedings +
Title Ranking related entities: Components and analyses +
Type conference paper  +
Year 2010 +
Creation dateThis property is a special property in this wiki. 8 November 2014 05:04:47  +
Categories Publications without license parameter  + , Publications without remote mirror parameter  + , Publications without archive mirror parameter  + , Publications without paywall mirror parameter  + , Conference papers  + , Publications without references parameter  + , Publications  +
Modification dateThis property is a special property in this wiki. 8 November 2014 05:04:47  +
DateThis property is a special property in this wiki. 2010  +
hide properties that link here 
Ranking related entities: Components and analyses + Title
 

 

Enter the name of the page to start browsing from.