About Me

Branislav Bédi

I am a PhD student in Second Language Studies at the University of Iceland and a project manager at The Árni Magnússon Institute for Icelandic Studies. The focus of my research is on natural language, multimodal behaviour of human speakers in conversations, natural language corpus, creating conversational scenarios, design of conversational behaviour of Embodied Conversational Agents (ECAs), computer games for language learning, and CALL in general. Besides my research, I am also interested in Applied Linguistics and teaching Icelandic as a second or foreign language.

Together with the Center for Analysis and Design of Intelligent Agents group (CADIA) at Reykjavik University, I was working on developing the Icelandic language and culture training application Virtual Reykjavik. Within this project, I developed a theoretical model for multimodal clarification requests (CRs) in Icelandic that can be implemented into the conversational behaviour of ECAs in order for the agents to be able to speak to human users in a more natural way. I also collaborated on creating conversational scenarios for human-agent interaction and educational material including design of lessons in this application.

Together with the University of Cambridge, I have been collaborating in developing 3D applications for teaching English in augmented reality by using Google Cardboard and Google Street View, which falls into the category of mobile assisted language learning (MALL). 

Since December 2018, I have been working with researchers from the University of Geneva on developing an online content for teaching reading skills in Icelandic. The project LARA - the Learning and Reading Assistant already involves many researchers and languages, archaic and modern, in the world, and welcomes all those who are interested in expanding its content.

Currently, I work on usability and designing user interface strategies of educational applications for language learning and crowdsourcing. This project is part of the COST action CA16105 EnetCollect.