Semantic processing and description of multiword expressions in CLANES.

Abstract

This chapter describes the process followed by the ACTRES Research Group to handle MWEs and leverage NLG tools in the food and drink industry. This goal will be reached in different interwoven stages: we will explain the methods followed to extract, tag, and standardise the MWE-related data retrieved from the CLANES corpus, a multilayer English-Spanish comparable corpus of the food and drink domain with more than 1.8 million words (Sanjurjo-González, in this volume);