Dear SIGEDU Community,
We are excited to invite participants to LANGLEARN, the first shared task on Language Learning Development.
Over the past years, there has been a growing interest in exploiting Natural Language Processing tools and machine learning methods for studying language development, both in first (L1) and second language (L2) acquisition scenarios. These systems have been devised to characterize the properties of learners’ language and how it evolves over time, across modalities and stages of acquisition.
The LANGLEARN shared task invites interested participants to submit their models for automatic language development assessment. The systems will be challenged to predict the relative order of a pair of essays written by students at different stages of their learning process.
The LANGLEARN task is multilingual as we rely on the following two corpora: i) CItA, a longitudinal corpus of Italian essays written by first language (L1) students in the first and second year of lower secondary school. ii) COWS-L2H, a corpus of short essays written in Spanish by second language (L2) students of an American university.
The goal of this shared task is not only to address open issues of automatic language development assessment, but also to gather for the first time the community of researchers and industry professionals interested in the topic.
Training data are available on the task website and the evaluation windows will be open between 2nd and 19th May 2023. Systems will be presented in Parma, Italy (7th - 8th September 2023) during EVALITA 2023 conference.
The LANGLEARN organizing committee