Skip to main navigation Skip to search Skip to main content

From retrieval to generative models: a design-based research approach to developing a chatbot for argumentative writing

  • Kai Guo
  • , Danling Li
  • , Jian Wang
  • , Samuel Kai Wah Chu

Research output: Contribution to journal › Article › peer-review

2 Citations (Scopus)

Abstract

Chatbots have emerged as a valuable tool in the field of English as a foreign language (EFL) education, particularly following the release of ChatGPT in November 2022. Numerous studies have since investigated the potential of chatbots for enhancing writing skills. Our interest in chatbot-supported writing instruction predates this release, and the advent of ChatGPT has opened new avenues for our research. This Innovative Practice paper details our design-based research project, initiated in September 2021 and spanning approximately three years, during which we developed Argumate, a chatbot designed to facilitate EFL students' argumentative writing. This paper delineates our development process, tracing our initial use of retrieval models in the chatbot's development to our current application of generative models in refining Argumate. Overall, this paper provides valuable insights into the process of developing chatbots to support students' argumentative writing.

Original languageEnglish
JournalInnovation in Language Learning and Teaching
DOIs
Publication statusAccepted/In press - 2025

Keywords

  • argumentative writing
  • Chatbots
  • design-based research
  • generative models
  • retrieval models

Fingerprint

Dive into the research topics of 'From retrieval to generative models: a design-based research approach to developing a chatbot for argumentative writing'. Together they form a unique fingerprint.

Cite this