Konkuk University AI Student Team Won an International Competition

The ‘KU-NLP Team’ (advisor Hak-su Kim), a student team from the Department of Artificial Intelligence at Konkuk University, recently took first place in the ‘AI chatbot text generation’ category at the ‘DialDoc 2021’, an international competition of natural language parsing.

The title of the team's paper was ‘Document-Grounded Goal-Oriented Dialogue Systems on Pre-Trained Language Model with Diverse Input Representation’.

Konkuk University graduate students Bo-eun Kim, Do-haeng Lee (Department of Artificial Intelligence), undergraduate student Ye-jin Lee (Department of Applied Statistics) and Kangwon National University Ph.D. student Si-hyung Kim formed a team to participate in the competition.

The competition, held by the Association for Computational Linguistics (ACL), the best international academic group in the field of natural language processing, is an international competition that about 20 teams from prestigious universities and companies around the world participated this year. The task was to create an AI chatbot that would acquire knowledge on its own and generate appropriate responses.

This year’s competition, which has been held for about three months since the data was released in February, was divided into two sections of ‘Knowledge identification’ and ‘Text generation’.

“It was not just a simple daily conversation, but a study on a document-based conversation system that uses knowledge information to generate responses. I would like to continue the study utilizing the relationship between dialogue structure and knowledge documents,” said Do-haeng Lee, a graduate student at Konkuk University who won the award.

Another team member Bo-eun Kim said, “It was a good experience to apply various ideas while implementing a document-based conversation generation model.”

Reported by 'Smart City Today'