A reverse dictionary based on semantic analysis using wordnet

Abstract

In this research we present a new approach for reverse dictionary creation, one purely semantic. We focus on a semantic analysis of input phrases using semantic similarity measures to represent words as vectors in a semantic space previously created assisted by WordNet. Then, applying algebraic analysis we select a sample of candidate words which passes through a filtering process and a ranking phase. Finally, a predefined number of output target words are displayed. A test set of 50 input concepts was created in order to evaluate our system, comparing our experimental results against OneLook Reverse Dictionary to demonstrate that our system provides better results over current available implementations. © Springer-Verlag 2013.