How would learning a second language (L2) change if students had access to tailored learning materials that matched their own personal interests? The interdisciplinary research project Using AI to topically personalise grammar learning will explore this question. 

Clearly, providing every learner with interest-matching learning materials is challenging. Doing this well will require a body of L2 practice material that touches on a broad range of topics and interests, with a variety that cannot typically be covered in traditional workbooks or by teachers.  

To help overcome the challenge, we will employ an e-learning platform featuring Automatic Exercise Generation technology, which will enable us to create a large bank of materials. This platform makes it possible to automatically transform authentic texts (e.g., news articles, blog posts, movie reviews) into grammar exercises of different types. The documents we select for input will be guided by the results of a general interest survey with our target population, to increase the likelihood of offering exercises with personally relevant topics. 

We plan to investigate how making these personalised learning materials available to students in the form of exercises impacts their learning outcomes, as well as their learning motivation and emotional state. Participants will work with different versions of the e-learning platform on multiple occasions, and complete proficiency tests and report their affective experiences and motivation in the practice sessions. 

In completing this project, we hope to better understand the degree to which people benefit from L2 instruction that is tailored to their individual interests. In the process, we also aim to contribute an innovative AI-based methodology that makes it possible to efficiently create L2 learning material.