Evaluating large-scale knowledge resources across languages

Abstract

This paper presents an empirical evaluation in a multilingual scenario of the semantic knowledge present on publicly available large-scale knowledge resources. The study covers a wide range of manually and automatically derived large-scale knowledge resources for English and Spanish. In order to establish a fair and neutral comparison, the knowledge resources are evaluated using the same method on two Word Sense Disambiguation tasks (Senseval-3 English and Spanish Lexical Sample Tasks). First, this study empirically demonstrates that the combination of the knowledge contained in these resources surpass the most frequent sense classi er for English. Second, we also show that this large-scale topical knowledge acquired from one language can be successfully ported to other languages.