Combining similarities with regression based classifiers for Entity Linking at TAC 2010

نویسندگان

  • César de Pablo-Sánchez
  • Juan Perea
  • Paloma Martínez
چکیده

The UC3M team has sent three runs for each Entity Linking task proposed in the Knowledge Base Population (KBP) track at TAC 2010. The skeleton system presented in 2009 has evolved in 2010 by incorporating some new tools, new algorithms for candidate retrieval and feature extraction, and two new stages that use regression based classifiers for candidate filtering. These improvements have allowed the UC3M team overall values to almost reach the median values of all participants in the Entity Linking tasks.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

TCAR at TAC-KBP-2010

The TCAR team developed multiple information retrieval based entity linking systems in a matter of weeks for the TAC-KBP evaluation task. We focused primarily on developing entity linking algorithms that do not require Wikipedia text and correctly detect when a given entity does not exist in Wikipedia (NIL). Without using Wikipedia text, the system achieves an overall TAC 2010 score of 67 perce...

متن کامل

Document-level Entity Linking: CMCRC at TAC 2010

This paper describes the CMCRC systems entered in the TAC 2010 entity linking challenge. The best performing system we describe implements the document-level entity linking system from Cucerzan (2007), with several additions that exploit global information. Our implementation of Cucerzan’s method achieved a score of 74.9% in development experiments. Additional global information improves perfor...

متن کامل

The Effect of Transitive Closure on the Calibration of Logistic Regression for Entity Resolution

This paper describes a series of experiments in using logistic regression machine learning as a method for entity resolution. From these experiments the authors concluded that when a supervised ML algorithm is trained to classify a pair of entity references as linked or not linked pair, the evaluation of the model’s performance should take into account the transitive closure of its pairwise lin...

متن کامل

BUPTTeam Participation at TAC 2013 Entity Linking

This paper overviews BUPTTeam’s participation in the Entity Linking task. We tackle the problem of entity disambiguation for large collections of online pages. Specifically, Named Entity Disambiguation is the task to polish the ambiguities of named entities. Several methods have been proposed to solve this problem, but they are largely only focused on one disambiguation mean, Entity Linking or ...

متن کامل

Stanford-UBC Entity Linking at TAC-KBP

This paper describes the joint Stanford-UBC knowledge base population system for the entity linking task. We modified our 2009 approach, which was based on frequencies of Wikipedia back-links, providing a context-independent mapping from strings to possible Wikipedia titles. We then built on this foundation, taking into account named-entity recognition (NER) and coreference resolution informati...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2010