Loading [MathJax]/extensions/TeX/enclose.js
Emotion-Semantic-Aware Dual Contrastive Learning for Epistemic Emotion Identification of Learner-Generated Reviews in MOOCs | IEEE Journals & Magazine | IEEE Xplore

Emotion-Semantic-Aware Dual Contrastive Learning for Epistemic Emotion Identification of Learner-Generated Reviews in MOOCs

Publisher: IEEE

Abstract:

Identifying the epistemic emotions of learner-generated reviews in massive open online courses (MOOCs) can help instructors provide adaptive guidance and interventions fo...View more

Abstract:

Identifying the epistemic emotions of learner-generated reviews in massive open online courses (MOOCs) can help instructors provide adaptive guidance and interventions for learners. The epistemic emotion identification task is a fine-grained identification task that contains multiple categories of emotions arising during the learning process. Previous studies only consider emotional or semantic information within the review texts alone, which leads to insufficient feature representation. In addition, some categories of epistemic emotions are ambiguously distributed in feature space, making them hard to be distinguished. In this article, we present an emotion-semantic-aware dual contrastive learning (ES-DCL) approach to tackle these issues. In order to learn sufficient feature representation, implicit semantic features and human-interpretable emotional features are, respectively, extracted from two different views to form complementary emotional-semantic features. On this basis, by leveraging the experience of domain experts and the input emotional-semantic features, two types of contrastive losses (label contrastive loss and feature contrastive loss) are formulated. They are designed to train the discriminative distribution of emotional-semantic features in the sample space and to solve the anisotropy problem between different categories of epistemic emotions. The proposed ES-DCL is compared with 11 other baseline models on four different disciplinary MOOCs review datasets. Extensive experimental results show that our approach improves the performance of epistemic emotion identification, and significantly outperforms state-of-the-art deep learning-based methods in learning more discriminative sentence representations.
Published in: IEEE Transactions on Neural Networks and Learning Systems ( Volume: 35, Issue: 11, November 2024)
Page(s): 16464 - 16477
Date of Publication: 24 July 2023

ISSN Information:

PubMed ID: 37486839
Publisher: IEEE

Funding Agency:


I. Introduction

The advances in technologies have brought great innovations to the field of education, and massive open online courses (MOOCs) have gained momentum as an innovative way of increasing access and equity in education [1]. With the expansion of MOOCs, the number of learners involved is also increasing. Therefore, a large number of unstructured text data have been generated in the course review area, which can reflect learners’ emotional state and learning experience. Even though MOOCs are very popular, the high attrition rate has always been a concern of their owners [2]. In order to help instructors guide and intervene with learners’ emotions, so as to reduce the attrition of MOOCs, it is important to understand learner-generated reviews in the courses and emotion identification could be used as the first step for further adaptive interventions.

References

References is not available for this document.