Investigating the Interplay between Text Difficulty and Prerequisite Relation Identification in Educational Texts

Abstract

Prerequisite relations (PR) are fundamental in knowledge acquisition and the applications of Artificial Intelligence to distance learning, particularly with regard to personalized learning plans. The role of these relations is to specify the sequence of information acquisition necessary for understanding a target concept. Despite their significance, identifying PRs in educational texts is challenging, mainly due to the lack of systematic procedures for their identification on educational texts. This paper contributes to the ongoing research on PR identification by exploring the relationship between text difficulty, assessed across various linguistic properties and target audiences, and prerequisite relations. We conducted a crowd-based study on the novel task of prerequisite concept ordering. The study yielded preliminary yet valuable insights into the impact of text difficulty on the task. Such evidence sheds light on the need to account for the linguistic properties of texts when identifying PRs, thus advancing the field’s comprehension of PRs within the educational landscape. Ultimately, we hope that this work could foster novel linguistically-aware research on PR.