An important objective of the Crossreads project is to study the interaction of languages in use on the island of Sicily in antiquity. It is one which completing the linguistic annotation of the I.Sicily digital corpus will help us to achieve. But while multi-word expressions are useful for studying language interaction, their linguistic annotation is far from being straightforward. As Victoria Fendel explains, this has been an ongoing issue in the field of natural language processing for decades. Her new blog can be found here.
