A Combined Pattern-based and Distributional Approach for Automatic Hypernym Detection in Dutch.
G.Y.R. Schropp, Els Lefever, Véronique Hoste · Ghent University Academic Bibliography (Ghent University) · 2013
This paper proposes a two-step approach to find hypernym relations between pairs of noun phrases in Dutch text. We first apply a pattern-based approach that combines lexical and shallow syntactic information to extract a list of candidate hypernym pairs from the input text. In a second step, distributional similarity information is used to filter the obtained list of candidate pairs. Evaluation of the system shows encouraging results and reveals that the distributional information particularly helps to improve the precision for context dependent hypernym pairs. The proposed hypernym module is considered an important step in building a semantic structure for automatically extracted terminology. As our approach does not require external lexical resources, it can be applied for any given Dutch input text and is particularly well suited for domain and user specific text.