Measuring the similarity between semantic relations that hold among entities is an important and necessary step in various Web related tasks such as relation extraction, informati...
Terminologies and other knowledge resources are widely used to aid entity recognition in specialist domain texts. As well as providing lexicons of specialist terms, linkage from t...
Angus Roberts, Robert Gaizasukas, Mark Hepple, Yik...
We present novel kernels based on structured and unstructured features for reranking the N-best hypotheses of conditional random fields (CRFs) applied to entity extraction. The fo...
Truc-Vien T. Nguyen, Alessandro Moschitti, Giusepp...
The named entity disambiguation task is to resolve the many-to-many correspondence between ambiguous names and the unique realworld entity. This task can be modeled as a classifi...
In this paper we describe Tokyo Institute of Technology’s attempt at the TREC2007 question answering (QA) track. Keeping the same theoretical QA model as for the TREC2006 task, ...
Edward W. D. Whittaker, Matthias H. Heie, Josef R....