论文信息 - UBC_UOS-TYPED: Regression for typed-similarity

UBC_UOS-TYPED: Regression for typed-similarity

We approach the typed-similarity task using a range of heuristics that rely on information from the appropriate metadata fields for each type of similarity. In addition we train a linear regressor for each type of similarity. The results indicate that the linear regression is key for good performance. Our best system was ranked third in the task.

Eneko Agirre | Mark Stevenson | Aitor Gonzalez-Agirre | German Rigau | Nikolaos Aletras

[1] Barbara Plank,et al. Proceedings of the Seventh conference on International Language Resources and Evaluation (LREC'10) , 2010 .

[2] Christiane Fellbaum,et al. Book Reviews: WordNet: An Electronic Lexical Database , 1999, CL.

[3] Michael I. Jordan,et al. Latent Dirichlet Allocation , 2001, J. Mach. Learn. Res..

[4] Mark Stevenson,et al. Computing Similarity between Cultural Heritage Items using Multimodal Features , 2012, LaTeCH@EACL.

[5] Eneko Agirre,et al. Personalizing PageRank for Word Sense Disambiguation , 2009, EACL.

[6] Mark Stevenson,et al. Computing similarity between items in a digital library of cultural heritage , 2013, JOCCH.

[7] Ian H. Witten,et al. The WEKA data mining software: an update , 2009, SKDD.

[8] Eneko Agirre,et al. Exploring Knowledge Bases for Similarity , 2010, LREC.

[9] Christopher D. Manning,et al. Incorporating Non-local Information into Information Extraction Systems by Gibbs Sampling , 2005, ACL.

[10] Ian H. Witten,et al. Learning to link with wikipedia , 2008, CIKM '08.

[11] Dan Klein,et al. Feature-Rich Part-of-Speech Tagging with a Cyclic Dependency Network , 2003, NAACL.

[12] David N. Milne. Computing Semantic Relatedness using Wikipedia Link Structure , 2007 .