Sentiment Analysis in Learning Management Systems Understanding Student Feedback at Scale Conference: ; ;

Mohammed Almutairi · 2025

Read Sentiment Analysis in Learning Management Systems Understanding Student Feedback at Scale Conference: ; ; by Mohammed Almutairi (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-05490.txt.

Introduction

Mohammed Almutairi explores, in a 2025 paper, adding sentiment analysis to online learning platforms. When classes moved online during the pandemic, teachers lost the facial cues that show how students feel, and written feedback alone can hide confusion or frustration.

The proposed system reads student comments for emotional tone at scale. A deep neural network using word embeddings, memory cells, and an attention step was trained for the task and compared with a simpler statistical baseline. The aim is to give teachers a live read on class mood, so they can adjust pace and support. The work frames emotion-aware feedback as one bridge across the online communication gap.

Catalog record

CREATORMohammed Almutairi
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-05490

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Sentiment Analysis in Learning Management Systems Understanding Student Feedback at Scale Conference: ; ; Mohammed Almutairi Abstract . During the wake of the Covid-19 pandemic, the educational paradigm has experienced a major change from in person learning traditional to online platforms. The change of learning convention has impacted the teacher - student especially in non-verbal communication. The absent of non-verbal communication has led to a reliance on verbal feedback which diminished the efficacy of the educational experience. This paper explores the integration of sentiment analysis into learning management systems (LMS) to bridge the student-teacher’s gap by offering an alternative approach to interpreting student feedback beyond its verbal context. The research involves data preparation, feature selection, and the development of a deep neural network model encompassing word embedding, LSTM, and attention mechanisms. This model is compared against a logistic regression baseline to evaluate its efficacy in understanding student feedback. The study aims to bridge the communication gap between instructors and students in online learning environments, offering insights into the emotional context of student feedback and ultimately improving the quality of online education. Figure 1. Bridging the Non-Verbal Communication Gap in Online Education through Sentiment Analysis Integration in Learning Management Systems 1. Introduction The education industry confronted unprecedented challenges during the spread of Covid-19, compelling millions of teachers and students to shift to online method of learning ( Cui et al. 2023 ) . In 2021, McKinsey highlighted that the proliferation of Covid-19 significantly accelerated the development of online educational platforms and courses ( Child et al. 2023 ) . Yet, it remains largely unexplored whether these newly developed platforms and courses truly cater to the optimal learning and interactions need of students. Both verbal and non-verbal interactions between students and instructors play an important role in enhancing the learning experience ( Zi Yi et al. 2023 ) . In traditional, in-person learning methods, instructors can leverage non-verbal indications, which often significantly influence teaching methods. For instance, instructors can perceive cues like eye contact that mirror confusion or misunderstanding through non-verbal communication. Expressions like narrowed eyes may indicate frustration, while wide-eyed surprise can suggest unexpected realizations ( Hodges 2023 ) . However, the transition to online learning methods has disrupted communications mode of interaction, specially sidelining non-verbal communication. Consequently, the educational journey has become predominantly reliant on students’ verbal feedback. The missingness of non-verbal interactions has expanded the gap between instructors and students, affecting the overall quality of the educational experience ( Güler 2023 ) . While sentiment analysis and opinion mining assist in monitoring feelings about learning contents from student feedback, and providing a deeper understanding beyond verbal communication ( NIRBHIK 2023 ) , there is existing gap in the literature regrading sentiment analysis during the learning feedback. Various studies have emphasized the importance of considering emotions expressed in feedback’s to better diagnose and identify non-verbal expressions ( Choudhary et al. 2023 ; Xu et al. 2023 ; Evans et al. 2023 ) . However, the exploration sentiment analysis in an education context remain underexplored. Despite the promising aspects of integrating sentiment analysis into Learning Management Systems (LMS), it remains unknown how learning experiences and educational efficacy will evolve to enhance student experiences. Using sentiment analysis during the learning journey provides a pathway to discern whether feelings and emotions in verbal communications are positive or negative. The implementation of this feature in the LMS can assist with understanding when we cannot observe face-to-face actions between instructors and students ( Manzon 2023 ) . This paper introduces a sentiment analysis model to be integrated into an LMS. The integration of sentiment analysis functionality aims to enhance the overall learning experience and support students by incorporating their feedback to refine and improve course content, thereby addressing the identified gap in the existing literature. In addressing the challenges of effectively interpreting student feedback in online learning environments, our initial step involved meticulous data preparation and preprocessing. This crucial phase ensured the consistency and integrity of our dataset, including the resolution of any missing data issues. We then embarked on a strategic process of featurization, where we carefully selected relevant features from the data and created new ones to enhance our analysis. Following this, we selected a logistic regression model as our baseline. This model choice was to establish a foundational benchmark against our model. we developed a sentiment analysis model utilizing a deep neural network. This model is composed of four integral components: a word embedding layer, Long Short-Term Memory (LSTM), an attention mechanism, and an output layer. The final stages of our work focused on the training, testing, and analysing of our model, ensuring its accuracy and efficacy in interpreting student feedback. 2. Method 2.1. Dataset We utilized the RateMyProfessor dataset, which captures students’ comments and ratings for educational courses. This dataset includes a ’star_rating’ feature that provides an overall student rating on a scale from 1 to 5. These ratings help classify student comments as positive, negative, or neutral. The ’comments’ feature contains valuable feedback about courses, teachers, and the educational environment. The RateMyProfessor dataset serves our purpose for several reasons: it offers a diverse range of student feedback on instructors and courses, thereby capturing various institutions and teaching styles. This is something not possible with a single LMS dataset alone. Additionally, This feedback can serve as a proxy for understanding broader student experiences in the academic environment. The dataset’s public availability also makes it a more accessible resource compared to many LMS platforms, where feedback is often restricted and private. Table 1. Attributes in RateMyProfessor Dataset ( He 2020 ) Attribute Description Professor name Name of the professor who is rated. School name University where the professor is currently teaching. Department name Department where the professor is currently working. Local name University’s locally known name. State name State in which the university is located. Year since first review The professor’s teaching age, from the first student evaluation to the time when analysis was done in 2019. Star rating The star rating of the professor’s overall quality. Points 3.5-5.0 are considered good, 2.5-3.4 is average, and 1.0-2.4 is poor according to RMP’s official standard. This star rating is the average score given to professors by all student comments. Take again Percentage of students who would choose this course again. Difficulty index Difficulty level of a course. Point 1 is easiest, and point 5 is hardest. The difficulty index is the average score given to professors by all students. Tags Tags that students chose to describe a professor. Post date Date when the student posted a course evaluation. Student star Star rating given to a professor by each student. Student-rated difficulty Difficulty index given to a professor by each student. Attendance Whether attendance for a course is mandatory or not. For credit Whether students chose a course for credit (yes or no). Would take again Whether students would choose a course again (yes or no). Grade Student’s final score of a course. Possible grades are A+, A, A-, B+, B, B-, C+, C, C-, D+, D, D-, F, WD, INC, Not, Audit/No. "WD" is Drop/Withdrawal, "INC" means Incomplete, "Not" is Not sure yet, and "Audit/No" is Audit/No Grade. Comment Comments that students gave

Read the complete text