I am co-organizing the third workshop on Computational Approaches on Linguistic Code-switching along with other researchers: Victor Soto (Columbia University), Julia Hirschberg (Columbia University), Fahad Ghamdi (George Washington University), Mona Diab (George Washington University), and Thamar Solorio (University of Houston).
The workshop will be collocated with ACL 2018 on July 19th, Melbourne, Australia. We will feature a shared task on Named Entity Recognition on Code-switched Twitter data for two language pairs: Spanish-English (SPA-ENG), and Modern Standard Arabic-Egyptian (MSA-EGY). Besides the code-switched behavior, we will also make the task more difficult than the traditional NER by adding more fine-grained entities; along with the standard person, location, and organization we have added product, title (movies, books, songs, etc.), group (sports teams, bands, etc.), event, time, and other. The idea is to encourage entity extraction by specific types that are actually treated differently. For instance, a movie (title) is not treated the same way as a soccer team (group) based on the context in which they are used. We expect that participants can use this contextual information to identify and recognize the entities.
We are excited to see 15 participants already registered on the shared task! If you are interested in participating, you can find the registration form here.
Hope to see you all in Australia!