Sentiment Analysis of Students’ Feedback with NLP and Deep Learning: A Systematic Mapping Study

Capuano N., Greco L., Ritrovato P., Vento M.

In recent years there has been a significant rethinking of corporate management, which is increasingly based on customer orientation principles. As a matter of fact, customer relationship management processes and systems are ever more popular and crucial to facing today’s business challenges. However, the large number of available customer communication stimuli coming from different (direct and indirect) channels, require automatic language processing techniques to help filter and qualify such stimuli, determine priorities, facilitate the routing of requests and reduce the response times. In this scenario, sentiment analysis plays an important role in measuring customer satisfaction, tracking consumer opinion, interacting with consumers and building customer loyalty. The research described in this paper proposes an approach based on Hierarchical Attention Networks for detecting the sentiment polarity of customer communications. Unlike other existing approaches, after initial training, the defined model can improve over time during system operation using the feedback provided by CRM operators thanks to an integrated incremental learning mechanism. The paper also describes the developed prototype as well as the dataset used for training the model which includes over 30.000 annotated items. The results of two experiments aimed at measuring classifier performance and validating the retraining mechanism are also presented and discussed. In particular, the classifier accuracy turned out to be better than that of other algorithms for the supported languages (macro-averaged f1-score of 0.89 and 0.79 for Italian and English respectively) and the retraining mechanism was able to improve the classification accuracy on new samples without degrading the overall system performance.

Zhou J., Ye J.

Sentiment analysis (SA) is widespread across all fields and has become one of the most active topics in education research, and there is a growing body of papers published. So far, however, there h...

Imran A.S., Daudpota S.M., Kastrati Z., Batra R.

How different cultures react and respond given a crisis is predominant in a society’s norms and political will to combat the situation. Often, the decisions made are necessitated by events, social pressure, or the need of the hour, which may not represent the nation’s will. While some are pleased with it, others might show resentment. Coronavirus (COVID-19) brought a mix of similar emotions from the nations towards the decisions taken by their respective governments. Social media was bombarded with posts containing both positive and negative sentiments on the COVID-19, pandemic, lockdown, and hashtags past couple of months. Despite geographically close, many neighboring countries reacted differently to one another. For instance, Denmark and Sweden, which share many similarities, stood poles apart on the decision taken by their respective governments. Yet, their nation’s support was mostly unanimous, unlike the South Asian neighboring countries where people showed a lot of anxiety and resentment. The purpose of this study is to analyze reaction of citizens from different cultures to the novel Coronavirus and people’s sentiment about subsequent actions taken by different countries. Deep long short-term memory (LSTM) models used for estimating the sentiment polarity and emotions from extracted tweets have been trained to achieve state-of-the-art accuracy on the sentiment140 dataset. The use of emoticons showed a unique and novel way of validating the supervised deep learning models on tweets extracted from Twitter.

Kandhro I.A., Jumani S.Z., Ali F., Shaikh Z.U., Arain M.A., Shaikh A.A.

This paper focuses on the performance analysis of hyperparameters of the Sentiment Analysis (SA) model of a course evaluation dataset. The performance was analyzed regarding hyperparameters such as activation, optimization, and regularization. In this paper, the activation functions used were adam, adagrad, nadam, adamax, and hard_sigmoid, the optimization functions were softmax, softplus, sigmoid, and relu, and the dropout values were 0.1, 0.2, 0.3, and 0.4. The results indicate that parameters adam and softmax with dropout value 2.0 are effective when compared to other combinations of the SA model. The experimental results reveal that the proposed model outperforms the state-of-the-art deep learning classifiers.

Chauhan P., Sharma N., Sikka G.

This work presents and assesses the power of various volumetric, sentiment, and social network approaches to predict crucial decisions from online social media platforms. The views of individuals play a vital role in the discovery of some critical decisions. Social media has become a well-known platform for voicing the feelings of the general population around the globe for almost decades. Sentiment analysis or opinion mining is a method that is used to mine the general population’s views or feelings. In this respect, the forecasting of election results is an application of sentiment analysis aimed at predicting the outcomes of an ongoing election by gauging the mood of the public through social media. This survey paper outlines the evaluation of sentiment analysis techniques and tries to edify the contribution of the researchers to predict election results through social media content. This paper also gives a review of studies that tried to infer the political stance of online users using social media platforms such as Facebook and Twitter. Besides, this paper highlights the research challenges associated with predicting election results and open issues related to sentiment analysis. Further, this paper also suggests some future directions in respective election prediction using social media content.

Barrón Estrada M.L., Zatarain Cabada R., Oramas Bustillos R., Graff M.

• Creation of two dataset for emotions and sentiments in text. • Recognitions of learning-centered emotions in text. • Comparison among machine & deep learning methods against an evolutionary approach. • Integration of best classification model to an intelligent learning environment. This paper presents a comparison among several sentiment analysis classifiers using three different techniques – machine learning, deep learning, and an evolutionary approach called EvoMSA – for the classification of educational opinions in an Intelligent Learning Environment called ILE-Java. To make this comparison, we develop two corpora of expressions into the programming languages domain, which reflect the emotional state of students regarding teachers, exams, homework, and academic projects, among others. A corpus called sentiTEXT has polarity (positive and negative) labels, while a corpus called eduSERE has positive and negative learning-centered emotions (engaged, excited, bored, and frustrated) labels. From the experiments carried out with the three techniques, we conclude that the evolutionary algorithm (EvoMSA) generated the best results with an accuracy of 93% for the corpus sentiTEXT, and 84% for the corpus eduSERE.

Kastrati Z., Imran A.S., Kurti A.

Students' feedback is an effective mechanism that provides valuable insights about teaching-learning process. Handling opinions of students expressed in reviews is a quite labour-intensive and tedious task as it is typically performed manually by the human intervention. While this task may be viable for small-scale courses that involve just a few students' feedback, it is unpractical for large-scale cases as it applies to online courses in general, and MOOCs, in particular. Therefore, to address this issue, we propose in this paper a framework to automatically analyzing opinions of students expressed in reviews. Specifically, the framework relies on aspect-level sentiment analysis and aims to automatically identify sentiment or opinion polarity expressed towards a given aspect related to the MOOC. The proposed framework takes advantage of weakly supervised annotation of MOOC-related aspects and propagates the weak supervision signal to effectively identify the aspect categories discussed in the unlabeled students' reviews. Consequently, it significantly reduces the need for manually annotated data which is the main bottleneck for all deep learning techniques. A large-scale real-world education dataset containing around 105k students' reviews collected from Coursera and a dataset comprising of 5989 students' feedback in traditional classroom settings are used to perform experiments. The experimental results indicate that our proposed framework attains inspiring performance with respect to both the aspect category identification and the aspect sentiment classification. Moreover, the results suggest that the framework leads to more accurate results than the expensive and labour-intensive sentiment analysis techniques relying heavily on manually labelled data.

Acheampong F.A., Wenyu C., Nunoo‐Mensah H.

Lundqvist K., Liyanagunawardena T., Starkey L.

Many course designers trying to evaluate the experience of participants in a MOOC will find it difficult to track and analyse the online actions and interactions of students because there may be thousands of learners enrolled in courses that sometimes last only a few weeks. This study explores the use of automated sentiment analysis in assessing student experience in a beginner computer programming MOOC. A dataset of more than 25,000 online posts made by participants during the course was analysed and compared to student feedback. The results were further analysed by grouping participants according to their prior knowledge of the subject: beginner, experienced, and unknown. In this study, the average sentiment expressed through online posts reflected the feedback statements. Beginners, the target group for the MOOC, were more positive about the course than experienced participants, largely due to the extra assistance they received. Many experienced participants had expected to learn about topics that were beyond the scope of the MOOC. The results suggest that MOOC designers should consider using sentiment analysis to evaluate student feedback and inform MOOC design.

Spatiotis N., Perikos I., Mporas I., Paraskevas M.

Learners’ opinions constitute an important source of information that can be useful to teachers and educational instructors in order to improve learning procedures and training activities. By analyzing learners’ actions and extracting data related to their learning behavior, educators can specify proper learning approaches to stimulate learners’ interest and contribute to constructive monitoring of learning progress during the course or to improve future courses. Learners-generated content and their feedback and comments can provide indicative information about the educational procedures that they attended and the training activities that they participated in. Educational systems must possess mechanisms to analyze learners’ comments and automatically specify their opinions and attitude towards the courses and the learning activities that are offered to them. This paper describes a Greek language sentiment analysis system that analyzes texts written in Greek language and generates feature vectors which together with classification algorithms give us the opportunity to classify Greek texts based on the personal opinion and the degree of satisfaction expressed. The sentiment analysis module has been integrated into the hybrid educational systems of the Greek school network that offers life-long learning courses. The module offers a wide range of possibilities to lecturers, policymakers and educational institutes that participate in the training procedure and offers life-long learning courses, to understand how their learners perceive learning activities and specify what aspects of the learning activities they liked and disliked. The experimental study show quite interesting results regarding the performance of the sentiment analysis methodology and the specification of users’ opinions and satisfaction. The feature analysis demonstrates interesting findings regarding the characteristics that provide indicative information for opinion analysis and embeddings combined with deep learning approaches yield satisfactory results.

Sangeetha K., Prabha D.

Classroom teaching becomes viable and efficient based on increase in participation of the student. This can be made possible by taking needed measure by finding the emotions of the students. Many researchers worked on emotion identification of students. Now-a-days sentiment analysis using deep learning models have gained good performance. Especially ensemble Long Short-Term Memory (LSTM) with attention layers gives more attention to the influence word on the emotion. In the proposed method, input sequences of sentences are processed parallel across multi-head attention layer with fine grained embeddings (Glove and Cove) and tested with different dropout rates to increase the accuracy. Later in this paper, the information from both deep multi-layers is fused and fed as input to the LSTM layer. In this paper, we conclude that the fusion of multiple layers accompanied with LSTM improves the result over a common Natural Language Processing method.

Nikolić N., Grljević O., Kovačević A.

Purpose Student recruitment and retention are important issues for all higher education institutions. Constant monitoring of student satisfaction levels is therefore crucial. Traditionally, students voice their opinions through official surveys organized by the universities. In addition to that, nowadays, social media and review websites such as “Rate my professors” are rich sources of opinions that should not be ignored. Automated mining of students’ opinions can be realized via aspect-based sentiment analysis (ABSA). ABSA s is a sub-discipline of natural language processing (NLP) that focusses on the identification of sentiments (negative, neutral, positive) and aspects (sentiment targets) in a sentence. The purpose of this paper is to introduce a system for ABSA of free text reviews expressed in student opinion surveys in the Serbian language. Sentiment analysis was carried out at the finest level of text granularity – the level of sentence segment (phrase and clause). Design/methodology/approach The presented system relies on NLP techniques, machine learning models, rules and dictionaries. The corpora collected and annotated for system development and evaluation comprise students’ reviews of teaching staff at the Faculty of Technical Sciences, University of Novi Sad, Serbia, and a corpus of publicly available reviews from the Serbian equivalent of the “Rate my professors” website. Findings The research results indicate that positive sentiment can successfully be identified with the F-measure of 0.83, while negative sentiment can be detected with the F-measure of 0.94. While the F-measure for the aspect’s range is between 0.49 and 0.89, depending on their frequency in the corpus. Furthermore, the authors have concluded that the quality of ABSA depends on the source of the reviews (official students’ surveys vs review websites). Practical implications The system for ABSA presented in this paper could improve the quality of service provided by the Serbian higher education institutions through a more effective search and summary of students’ opinions. For example, a particular educational institution could very easily find out which aspects of their service the students are not satisfied with and to which aspects of their service more attention should be directed. Originality/value To the best of the authors’ knowledge, this is the first study of ABSA carried out at the level of sentence segment for the Serbian language. The methodology and findings presented in this paper provide a much-needed bases for further work on sentiment analysis for the Serbian language that is well under-resourced and under-researched in this area.

Hew K.F., Hu X., Qiao C., Tang Y.

This study defines MOOC success as the extent of student satisfaction with the course. Having more satisfied MOOC students can extend the reach of an institution to more people, build the brand name of the institution, and even help the institution use MOOCs as a source of revenue. Traditionally, student completion rate is frequently used to define MOOC success, which however, is often inaccurate because many students have no intention of finishing a MOOC. Informed by Moore's theory of transactional distance, this study adopted supervised machine learning algorithm, sentiment analysis and hierarchical linear modelling to analyze the course features of 249 randomly sampled MOOCs and 6393 students' perceptions of these MOOCs. The results showed that course instructor, content, assessment, and schedule play significant roles in explaining student satisfaction, while course structure, major, duration, video, interaction, perceived course workload and perceived difficulty play no significant roles. This study adds to the extant literature by examining specific learner-level and course-level factors that can predict MOOC learner satisfaction and estimating their relative effects. Implications for MOOC instructors and practitioners are also provided. • Examines learner-level and course-level factors that can predict MOOC learner satisfaction. • Examines 249 randomly sampled MOOCs, comprising 6393 students. • Course instructor, content, assessment, and schedule significantly predict student satisfaction. • Course major, duration, perceived workload and perceived difficulty play no significant roles.

Yang L., Li Y., Wang J., Sherratt R.S.

In recent years, with the rapid development of Internet technology, online shopping has become a mainstream way for users to purchase and consume. Sentiment analysis of a large number of user reviews on e-commerce platforms can effectively improve user satisfaction. This paper proposes a new sentiment analysis model-SLCABG, which is based on the sentiment lexicon and combines Convolutional Neural Network (CNN) and attention-based Bidirectional Gated Recurrent Unit (BiGRU). In terms of methods, the SLCABG model combines the advantages of sentiment lexicon and deep learning technology, and overcomes the shortcomings of existing sentiment analysis model of product reviews. The SLCABG model combines the advantages of the sentiment lexicon and deep learning techniques. First, the sentiment lexicon is used to enhance the sentiment features in the reviews. Then the CNN and the Gated Recurrent Unit (GRU) network are used to extract the main sentiment features and context features in the reviews and use the attention mechanism to weight. And finally classify the weighted sentiment features. In terms of data, this paper crawls and cleans the real book evaluation of dangdang.com, a famous Chinese e-commerce website, for training and testing, all of which are based on Chinese. The scale of the data has reached 100000 orders of magnitude, which can be widely used in the field of Chinese sentiment analysis. The experimental results show that the model can effectively improve the performance of text sentiment analysis.

Shuang K., Zhang Z., Loo J., Su S.

Existing unsupervised word embedding methods have been proved to be effective to capture latent semantic information on various tasks of Natural Language Processing (NLP). However, existing word representation methods are incapable of tackling both the polysemous-unaware and task-unaware problems that are common phenomena in NLP tasks. In this work, we present a novel Convolution–Deconvolution Word Embedding (CDWE), an end-to-end multi-prototype fusion embedding that fuses context-specific information and task-specific information. To the best of our knowledge, we are the first to extend deconvolution (e.g. convolution transpose), which has been widely used in computer vision, to word embedding generation. We empirically demonstrate the efficiency and generalization ability of CDWE by applying it to two representative tasks in NLP: text classification and machine translation. The models of CDWE significantly outperform the baselines and achieve state-of-the-art results on both tasks. To validate the efficiency of CDWE further, we demonstrate how CDWE solves the polysemous-unaware and task-unaware problems via analyzing the Text Deconvolution Saliency, which is an existing strategy for evaluating the outputs of deconvolution.

Ramezani E.B.

Awadh W.A., Sulaiman R.B., Mahmoud M.A.

Ashfaq F., Jhanjhi N.Z., Khan N.A., Muzafar S., Das S.R.

Chedrawi C., Raya K., Kazoun N.

Purpose This paper aims to study social media (SM) role in detecting food supply chain management (SCM) challenges and the importance of the Blockchain technology (BCT) in presenting sustainable solutions to these challenges. Design/methodology/approach This paper follows a big data analytics approach by using text mining techniques and raw data from X/Twitter and processed using the open-source programming language R. Findings This paper shows that SM generally and X/Twitter particularly play a major role in detecting FSC challenges and in shedding the light on trends and gaps in food supply chain (SC), along with the main actors on SM that influence food SCM. This study also shows that BCT is a convenient technology to reconsider the structure of food SCM to ensure higher food safety and a better digitalization of food SC, especially in the Asian region, with the large population and key consumers in food market. Originality/value This study contributes to data and knowledge that try to fill gaps in research related to the role of SM in detecting food SCM challenges through BCT.

Rana M.R., Nawaz A., Rehman S.U., Abid M.A., Garayevi M., Kajanová J.

Sentiment analysis plays an important role in understanding employee feedback and improving workplace culture. By leveraging NLP techniques to analyze this feedback accurately, organizations can pinpoint specific areas that need improvement, address employee concerns, and foster a positive work environment. These NLP-driven deep learning models offer valuable tools for E-Commerce HR and sales departments, enabling monitoring employee and users’ sentiment trends over time and assisting in implementing targeted interventions. Focusing on the e-commerce industry, this work utilizes NLP-driven deep learning methodologies to analyze employee and user feedback, aiming to identify sentiments. The proposed NLP-driven, deep learning-based framework is designed to classify user feedback into positive, negative, or neutral sentiments. The key steps in this framework include data collection, NLP-enhanced feature extraction using BERT-BiGRU, and final classification using a Graph Neural Network-based finite-state automata. The effectiveness of this NLP-centric approach was tested on diverse datasets of customer feedback from the e-commerce industry. The results demonstrate the framework’s efficacy, achieving an impressive 93.35% accuracy rate, surpassing existing benchmark methods. The research significantly benefits e-commerce by refining product portfolios and enhancing workplace culture.

A S S., Nagasundaram S.

Devi S.A., Ram M.S., Dileep P., Pappu S.R., Rao T.S., Malyadri M.

Talele A., Barhate M., Chakrabarty S., Jain S., Bhat S., Bisen S.

Kothuri S.R., Rajalakshmi N.R.

Multimodal Sentiment Analysis (MSA) is a growing area of emotional computing that involves analyzing data from three different modalities. Gathering data from Multimodal Sentiment analysis in Car Reviews (MuSe-CaR) is challenging due to data imbalance across modalities. To address this, an effective data augmentation approach is proposed by combining dynamic synthetic minority oversampling with a multimodal elicitation conditional generative adversarial network for emotion recognition using audio, text, and visual data. The balanced data is then fed into a granular elastic-net regression with a hybrid feature selection method based on dandelion fick’s law optimization to analyze sentiments. The selected features are input into a multilabel wavelet convolutional neural network to classify emotion states accurately. The proposed approach, implemented in python, outperforms existing methods in terms of trustworthiness (0.695), arousal (0.723), and valence (0.6245) on the car review dataset. Additionally, the feature selection method achieves high accuracy (99.65%), recall (99.45%), and precision (99.66%). This demonstrates the effectiveness of the proposed MSA approach, even with three modalities of data.

Mayormente M.D., Gumpal B.R.

Bensoltane R., Zaki T.

Most existing aspect-based sentiment analysis (ABSA) methods perform the tasks of aspect extraction and sentiment classification independently, assuming that the aspect terms are already determined when handling the aspect sentiment classification task. However, such settings are neither practical nor appropriate in real-life applications, as aspects must be extracted prior to sentiment classification. This study aims to overcome this shortcoming by jointly identifying aspect terms and the corresponding sentiments using a multi-task learning approach based on a unified tagging scheme. The proposed model uses the Bidirectional Encoder Representations from Transformers (BERT) model to produce the input representations, followed by a Bidirectional Gated Recurrent Unit (BiGRU) layer for further contextual and semantic coding. An attention layer is added on top of BiGRU to force the model to focus on the important parts of the sentence. Finally, a Conditional Random Fields (CRF) layer is used to handle inter-label dependencies. Experiments conducted on a reference Arabic hotel dataset show that the proposed model significantly outperforms the baseline and related work models.

Jazuli A., Widowati, Kusumaningrum R.

Evaluating the learning process requires a platform for students to express feedback and suggestions openly through online reviews. Sentiment analysis is often used to analyze review texts but typically captures only overall sentiment without identifying specific aspects. This study develops an aspect-based sentiment analysis (ABSA) model using IndoBERT, a pre-trained model tailored for the Indonesian language. The research uses 10,000 student reviews from Indonesian universities, processed through data labeling, text preprocessing, and splitting, followed by model training and performance evaluation. The model demonstrated superior performance with an aspect extraction accuracy of 0.973, an F1-score of 0.952, a sentiment classification accuracy of 0.979, and an F1-score of 0.974. Experimental results indicate that the proposed ABSA model surpasses previous state-of-the-art models in analyzing sentiment related to specific aspects of educational evaluation. By leveraging IndoBERT, the model effectively handles linguistic complexities and provides detailed insights into student experiences. These findings highlight the potential of the ABSA model in enhancing learning evaluations by offering precise, aspect-focused feedback, contributing to strategies for improving the quality of higher education.

Labib L.N., ElSabry E.A.

In the rapidly evolving landscape of technology, innovation, and sustainability, integrating Artificial Intelligence (AI) in higher education presents significant opportunities to enhance learning, teaching and assessment, streamline administrative processes, and promote sustainable educational practices. Despite a vast amount of literature on AI in education, comprehensive use cases that can inform effective implementation of AI in higher education remain scarce. This chapter aims to fill this gap by exploring in-depth the AI applications in higher education, addressing key areas of AI integration such as curriculum design and content development, pedagogical strategies and learning environments, evaluation mechanisms and feedback systems, process streamlining and efficiency enhancement, learning analytics, academic research and ideation, and student support and services. As such, this chapter contributes to the broader goal of sustainable development in the context of education thus aligning with the themes of the SMART conference. This research may inform sustainable solutions, policy, and governance mechanisms in the context of AI-driven education.

Mamani-Coaquira Y., Villanueva E.

Malashin I.P., Tynchenko V.S., Gantimurov A.P., Neluyb V.A., Borodulin A.S.