Abstract

Well trained linguists manage to capture semantic behavior of words in various annotated corpora. Using them as training data, semantic relations can be discovered by intelligent systems using supervised machine learning techniques. What if we have short deadlines and limited human and financial possibilities that prevent us from building such a valuable training corpus for our own language? If such a corpus already exists for any other language, we could make use of this treasure and reproduce it for the language we need. This paper proposes an import method, which transfers semantic annotation (which could be semantic roles, named entity, sentiments, etc.) from an annotated resource to another language, using comparable texts. The case of semantic role annotation transfer from English to Romanian is discussed.

Recommended Citation

Trandabăț, D.M. & Gagea, O. (2016). A New Perspective on Reusing Semantic Resources. In J. Gołuchowski, M. Pańkowska, C. Barry, M. Lang, H. Linger, & C. Schneider (Eds.), Information Systems Development: Complexity in Information Systems Development (ISD2016 Proceedings). Katowice, Poland: University of Economics in Katowice. ISBN: 978-83-7875-307-0. http://aisel.aisnet.org/isd2014/proceedings2016/CogScience/1.

Paper Type

Event

Share

COinS
 

A New Perspective on Reusing Semantic Resources

Well trained linguists manage to capture semantic behavior of words in various annotated corpora. Using them as training data, semantic relations can be discovered by intelligent systems using supervised machine learning techniques. What if we have short deadlines and limited human and financial possibilities that prevent us from building such a valuable training corpus for our own language? If such a corpus already exists for any other language, we could make use of this treasure and reproduce it for the language we need. This paper proposes an import method, which transfers semantic annotation (which could be semantic roles, named entity, sentiments, etc.) from an annotated resource to another language, using comparable texts. The case of semantic role annotation transfer from English to Romanian is discussed.