A distributional semantics model for idiom detection the case of english and Russian

Abstract

This paper describes experiments in English and Russian automatic idiom detection. Our algorithm is based on the idea that literal and idiomatic expressions appear in different contexts. This difference is captured by our distributional semantics model. We evaluate our model on both languages and compare its results. We show that our model is language-independent. We also describe a new annotated resource we created for our experiments.