(CLT) Data-driven learning meets generative artificial intelligence: Informal internet written language in Chinese, Japanese, and Korean

REGISTER HERE (@hawaii.edu required)

Facilitators: UH Mānoa, Center for Language & Technology
Alexander F. Tang, CLT Graduate Assistant
Keisuke Harada, EALL PhD student

Advisor: Dr. Richard Medina, CLT Associate Faculty Specialist in Human-Computer Interaction

Date: October 7, 2026 (Wednesday)

Time: 12:00PM to 1:00PM HST

Two ways to participate:

• In-Person: Moore Hall 258, UH Mānoa
• Zoom: Receive the link upon registering

Description: How can language teachers combine corpus-based language pedagogy (CBLP), data-driven learning (DDL), and generative artificial intelligence (GenAI) to create meaningful learning experiences beyond the English language classroom? In this professional learning session, we showcase the development of a bespoke Google Gemini chatbot (Google Gem) designed to help learners explore authentic informal internet language in Chinese, Japanese, and Korean. Drawing on corpora created from YouTube comments on mukbang videos, the chatbot uses a retrieval-augmented generation (RAG) approach to allow learners to investigate collocations, concordance examples, frequency patterns, and other features of naturally occurring online language. We will demonstrate our design process and discuss how corpus-informed GenAI tools can extend principles of CBLP by making DDL more accessible, interactive, and adaptable for learners and teachers of less commonly taught languages.

We will also share examples from an ongoing study of 60 learners of Chinese, Japanese, and Korean who used the chatbot alongside DDL activities to notice, compare, question, monitor, and evaluate authentic language use and GenAI-generated responses. Rather than positioning GenAI as a replacement for learner inquiry or teacher guidance, we consider how bespoke chatbots can support CBLP by facilitating critical engagement with language data while giving teachers greater control over the corpora and materials that inform AI responses. This session is intended for language teachers, teacher educators, and researchers interested in corpus linguistics, CBLP, DDL, CALL, or GenAI, including those with little or no prior experience creating chatbots. As this session also serves as a practice talk for the Symposium on Second Language Writing, we especially welcome feedback and discussion that can help us strengthen the presentation and consider how this approach might be adapted across languages and instructional contexts.