DIRECT: Toward Dialogue-Based Reading Comprehension Tutoring

Abstract

A major challenge in education is to provide students with a personalized learning experience. This study aims to address this by developing a dialogue-based intelligent tutoring system (ITS) that imitates human expert tutors. The ITS asks questions, assesses student answers, provides hints, and even chats to encourage student engagement. We constructed the Dialogue-based Reading Comprehension Tutoring (DIRECT) dataset to simulate real-world pedagogical scenarios with the assessment labels and key sentences to support tutoring. The DIRECT dataset is based on RACE, which is a large-scale English reading comprehension dataset. In addition, we propose a neural pipeline approach to model the tutoring tasks and conduct a comprehensive analysis on the results, including a human evaluation. The results show that our model performs well in generating questions, assessing answers, and chatting, showing high potential although some challenges remain. The proposed model provides a good basis for further development of dialogue-based ITSs.INDEX TERMS Computer aided instruction, dialogue-based tutoring, educational technology, intelligent tutoring, natural language processing.

Empowering Personalized Learning through a Conversation-based Tutoring System with Student Modeling

“…Recent advances in conversation-based tutoring systems utilizing LLM's have showcased their transformative potential [18]. For instance, ArgueTutor [32] offers adaptive tutoring for argumentation skills, DIRECT [16] specializes in reading comprehension tutoring systems, MATHDIAL [20] focuses on math reasoning problems, and the work by [27] enhances language learning experiences. Moreover, prompt engineering-based tutoring systems have broadened the horizons of possibilities [4,14].…”

“…Accordingly, a growing number of educators are showing interest in leveraging these LLM's to design an all-time available, personalized tutoring system 1 [18,21]. For example, prior works have developed LLM-driven conversation-based tutors in various skill domains, including argumentation [32], reading comprehension [16], mathematical reasoning [20], and language learning [27]. However, most prior works lack a systematic student assessment, relying completely on text inputs implicitly.…”