Wikipedia Sets: Context-Oriented Related Entity Acquisition from Multiple Words
Masumi Shirakawa, Kotaro Nakayama, Takahiro Hara, Shojiro Nishio · 2011
In this paper, we propose a method which acquires related words (entities) from multiple words by naturally disambiguating their meaning and considering their contexts. In addition, we introduce a bootstrapping method for improving the coverage of association relations. Experimental result shows that our method can acquire related words depending on the contexts of multiple words compared to the ESA-based method.