A Review of Sentiment Analysis Approaches for Quality Assurance in Teaching and Learning

doi:10.21203/rs.3.rs-2467791/v1

Download PDF

Short Report

A Review of Sentiment Analysis Approaches for Quality Assurance in Teaching and Learning

https://doi.org/10.21203/rs.3.rs-2467791/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The education industry considers quality to be a crucial factor in its development. Yet, the quality of many institutions is far from perfect as there is high rate of systemic failure and low performance among students. Consequently, the application of digital computing plays an increasingly important role in assuring the overall quality of an educational institution. However, literature lacks a reasonable number of systematic review that classifies researches that applied natural language processing and machine learning solutions for students’ feedback in sentiment analysis and quality assurance. Thus, this paper presents a systematic literature review that structure available published papers between 2014 and 2023 in high impact journal indexed databases. The work extracted 59 relevant papers out of the 3392 that was initially found using an exclusion and inclusion criteria. The result identified five (5) prevalent techniques that are majorly researched for sentiment analysis in the area of education as well as the prevalent supervised machine learning algorithms, lexicon-based approaches and evaluation metrics in assessing feedbacks in the education domain.

quality assurance

opinion mining

sentiment analysis

machine learning algorithms

The education system presently represents a landscape that is enriched by a continuous massive amount of data which are generated in different format on a daily basis, embedded in this data are valuable and useful information. Discovering and extracting this information from the large amount of data is one of the benefits that opinion mining and sentiment analysis can give. Opinions and sentiments that are being expressed by students are valuable information that can be used for analyzing the opinion of students about teachers, courses and topics. Though opinion mining and sentiment analysis appear similar, they slightly vary from each other. Opinion meaning refers to the process of extracting and analyzing individuals’ opinion about a particular subject, while sentiments analysis refers to the process of finding sentiment words or phrases that exhibits emotion. In this paper, we used both techniques interchangeably, sentiment/opinion polarity (positive, negative or neutral) signifies someone opinion towards a subject, while emotions represent someone’s feelings towards a subject. In this paper, we presented a systematic review of sentiment analysis on students’ feedback. The aim is to evaluate and present a general summary of research findings and implications for the purpose of research and practice. This is needed to provide updates concerning state of research, identification of well researched areas, revealing lagging areas that needs further research, and understanding similar challenges that occur.

The remaining part of this paper are in the following order; “Background and related work” which gives information on the application of how sentiment analysis has been used on students’ feedback. “Research methods” which discussed the adopted research methodology. “Result” which shows the findings of the study. The “Identified gaps and challenges” section presents the challenges in the reviewed papers on sentiment analysis of student’ feedback. The “limitation of the review” section shows the limitation of the study, while the paper is concluded in the “conclusion and future work” section.

Background

Lots of theories regarding emotion detection and analysis have been established since 1960s. The study conducted by [1] grouped emotions into eight groups which are joy, anticipation, anger, disgust, fear, trust, surprise, and sadness. Documents, sentences and words are different levels in which sentiment analysis can be carried out. But, due to documents, handling sentiment manually is imperial, for this reason, an automatic data processing is required. Natural Language Processing (NLP) can be used on text-based sentiment analysis or document level corporal. The majority of studies identified in the research up to 2016–2017 used only NLP methods, such as sentiment analytical techniques based on lexicon and dictionaries. Those papers actually rarely made use of traditional machine learning classifiers. Both recognition and classification of sentiment has recently undergone a change from purely NLP-based techniques to deep learning-based models, and the number of papers recently published on the study issue has greatly grown. Recently, both the popularity and relevance of student feedback have risen, especially during COVID-19 pandemic when most educational institutions shifted from traditional face-to-face interaction to an online format. The amount of new research indicates that there is a growing interest in using NLP, or machine learning techniques towards sentiment analysis in the area of education. To the best of our knowledge, the literature body lacks a review that systematically classifies and categorizes research and outcomes by showing the frequencies and summaries of publications and trends in order to determine the state of evidence in education. In order to carry out a systematic review, this article uses a process structure to respond to research questions. In particular, we created a number of research questions that address general concerns about the researched sentiment analysis elements, models, methodologies, and trends in assessment metrics in the teaching and learning community.

Related work

According to past studies, we observed that one study [2] on sentiment analysis (SA) in the field of education, concentrated on identifying the methodologies and tools utilized in SA as well as the major importance of using SA on educational data. Our study is an expanded version of this research, therefore a lot of data from different sources, including bibliographic sources, research trends and patterns, and the most current SA tools is provided. A summary of sentiment analysis techniques for education was presented in a review study by [3]. For multimodal fusions, the authors of this study presented a sentiment detection and assessment framework. Our review paper seeks to cover all issues related to the sentiment analysis of educational content focusing on textual information in a systematic way as opposed to the text, audio, and visual signals focused in [3]. Additionally, we provide a detailed review of current approaches used for sentiment discovery along with the results they achieved. Similar to [4], which reviewed the research journals of SA on education data and helped identify areas for further study, the writers of [4] covers subjects like the building of sentiment analysis systems, the examination of topics that are relevant to students, the analysis of teachers' teaching ability, etc., from about 41 related published research. In contrast, we first screened 618 research papers from various publications and conferences before conducting our scientific literature review analysis. We finalized and incorporated 59 of the most relevant and excellent scientific publications published from 2014 to 2023 in this study. The major goal of this work is to systematically compile all of the material currently available on sentiment analysis of educational data in one place. Such review studies are very beneficial for readers in this domain. This review study will help researchers, academicians and practitioners who are interested in sentiment analysis and quality assurance in the education domain.

The method that was adopted in this study is a systematic literature review of tools and technologies used in analyzing student opinion in higher education by adopting [5] and [6] as models.

Research question

The research questions (RQs) devised for this study were as follows:

RQ1. What are the most explored aspects in the area of education with regard to sentiment analysis?
RQ2. Which techniques and models are extensively researched for using sentiment analysis in the area of education?
RQ3. What are the most common metrics for measuring the effectiveness of sentiment analysis systems?
RQ4. What are the most popular methods for gathering student feedback?

Search string

To create a good search string, you need to structure your keyword phrase in terms of the comparison, intervention, population, and outcome [5]. Relevant papers were obtained by constructing a search phrase using keywords based on the previously stated research question. Seven (7) common database indexes which are Scopus, EBSCOhost, Science Direct, IEEE Xplore, Web Science, SpringerLink, and ACM DL were used to carry out the searches. The search strings are eleven (11) in total; they are “sentiment analysis”, “opinion mining”, “technologies used in sentiment analysis”, “sentiment analysis framework”, “sentiment analysis algorithms”, “sentiment analysis tools”, “students’ feedback”, “teacher assessment”, “feedback assessment”, “learners’ feedback sentiment analysis reviews” and “quality assurance”.

Data sources

It is more practical to choose from broad and standardized databases as research gets more multidisciplinary, international, and interactive. The following databases were consulted:

Scopus: Scopus is a database which was launched in 2004 and it includes citations and abstracts for academic journal articles. It provides a thorough picture of the world's research output in the scientific, technical, medical, and social disciplines and contains over 36,377 publications from over 11,678 publishers. It is the biggest database of peer-reviewed literature citations and abstracts.
ScienceDirect: This database is Elsevier top information resource for students and information professionals. It offers both open access and subscriber access to a sizable database that combines credible, proper scientific, technical, and healthcare papers with clever, user-friendly features. It has more than 35,000 books and over 14,000,000 publications from over 3,800 journals.
EBSCO: Researchers can access a variety of comprehensive and bibliographic databases through EBSCOhost, which again offers digital journal services for both academic and corporate researchers. Over 900,000 high-quality e-books and publications, 16,711 indexed journals, 14,914 of which come from peer-reviewed sources, over 60,000 recordings, and more than 1500 prominent academic publishers are all included.
IEEE Xplore: This database is a research resource for finding and accessing conference proceedings, journal articles and documented relating to computer science, electronics and electrical engineering. IEEE Xplore has over 300 peer-reviewed journals, 1,900 international conferences, over 11,000 technical standards, approximately 5,000 e-books, and more than 500 online courses
Web Science: This platform formerly known as Web of Knowledge is a platform with paid subscription that gives users access to several databases with reference and citation information from conference proceedings, academic journals, and other publications in a range of academic subjects.
SpringerLink: This database is the largest online library of books, journals, series, protocols and reference materials for science and technology. The database provides millions of scientific documents to researchers.
ACM DL: The ACM DL is a database for research discovery which contain Full-Text collection of publications including books, journals, conference proceedings technical magazines, and newsletters.

Data retrieval

The majority of high impact journals and conferences are indexed in this collection of comprehensive databases. The eleven (11) search words were joined using Boolean 'OR'. As displayed in Table 1, a total of 3,392 articles from the seven databases were retrieved.

Table 1: First search string result

	Scopus	ScienceDirect	EBSCOhost	IEEE Xplore	Web Science	SpringerLink	ACM DL	Total
Number of papers	821	437	681	576	465	268	144	3,392

The search was further streamlined by restricting it to computer science related papers and papers published between 2004 and 2023. At this point, 618 papers remained after a total of 2,774 papers were removed as displayed in Table 2

Table 2: Second search string result

	Scopus	ScienceDirect	EBSCOhost	IEEE Xplore	Web Science	SpringerLink	ACM DL	Total
Number of papers	129	89	92	99	82	73	54	618

After the second search, we went through the titles of the 618 remaining papers and discovered that only 292 have relevant titles as shown in Table 3

Table 3: Papers with relevant titles

	Scopus	ScienceDirect	EBSCOhost	IEEE Xplore	Web Science	SpringerLink	ACM DL	Total
Number of papers	69	44	51	34	41	31	22	292

Next, we went through the abstracts and introduction of the papers with relevant titles to know if they are at variance with our research questions that has earlier been stated. To facilitate analysis, the papers’ citations were exported to Microsoft Excel and three categories were used to classify the papers, these categories are “relevant”, “partially relevant” and “not relevant”. The relevant papers were marked with green color; the partially relevant papers were marked with yellow colors while the not relevant papers were marked with red color. At this point, 88 papers were determined to be “relevant,” 74 papers to be “partially relevant,” and 130 papers to be “irrelevant”. Examining the abstract carefully once more, 233 papers were discarded using the exclusion criteria, leaving a total of 59 papers as shown in Table 4, which underwent qualitative evaluation in accordance with the research questions.

Table 4: Final selection result

	Scopus	ScienceDirect	EBSCOhost	IEEE Xplore	Web Science	SpringerLink	ACM DL	Total
Number of papers	13	10	11	6	8	7	5	59

Eligibility criteria

Inclusion criteria

Papers from peer-reviewed conferences, journals, workshops and are also between the year 2014 and 2023 were included. Additionally, in cases where there were publications with identical studies and outcomes, the most current papers were chosen.

Exclusion criteria

Papers that are not written in English language, not relate to sentiment analysis, and whose contributions to the work are not explicitly stated in the abstract were all excluded from the reviewed papers in this study.

The study's results are now presented in relation to the research questions that guided the conduct of the systematic literature review.

RQ1. What are the most investigated aspects in the education domain with respect to sentiment analysis?

The opinions of students are a useful instrument for gaining important knowledge on different educational entities, such as lecturers, institutions, classes, etc. and teaching approaches involving these entities. It is crucial to recognize these aspects as they are expressed in students' textual remarks because it helps decision-makers take the necessary steps to address them specifically. In this context, we looked at and categorized the reviewed articles according to the issues that the authors wanted to look into and that were relevant to students. Specifically, we discovered three groups and associated teaching aspects that were the focus of these studies' research. The first group of research looked at how students responded to different qualities of their teachers, such as their knowledge, behavior, pedagogy, etc. The second group includes publications addressing other facets of the three distinct entities, such as courses, teachers, and institutions. Tuition costs, the campus, student life, and other characteristics connected to the institution entity were examples of course-related features. Course-related aspects comprised dimensions like course content, course structure, evaluation, etc. meanwhile the third group includes Papers examining the perspectives and attitudes of students toward institutional entities. From our findings as illustrated in Table 5, we found out that 76% of the papers that were reviewed were based on extracting students’ thoughts, opinion and attitudes toward teachers and 16% were based on extracting students’ opinion towards courses and institution, whereas the remaining 8% were based on extraction student opinion towards institution.

Table 5: Student Feedback Aspects Examined in the Reviewed Papers

Students’ opinion	Towards Teacher	Towards Institutions	Towards courses and institutions
Percentage	76%	8%	16%

RQ2. Which techniques and models are extensively researched for using sentiment analysis in the area of education?

A wide variety of techniques and models have been used to conduct sentiment analysis. These techniques are generally classified into three group which are supervised learning, unsupervised learning and lexicon-based techniques. While some researchers decide to use either supervised, unsupervised or the lexicon-based techniques, other researcher decide to use a hybrid of two of the three major techniques. Table 6 shows the learning techniques used for sentiment analysis in the area of education

Table 6: Learning techniques used for sentiment analysis in education domain

Learning Techniques	Papers
Supervised	[7], [8], [9], [10], [11], [12], [13], [2], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29].
Unsupervised	[30], [31], [32], [33], [34].
Lexicon-based	[35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47].
Supervised and unsupervised	[48], [49], [50], [51].
Supervised and lexicon-based	[52], [53], [54], [55], [56], [57], [58], [59], [60], [61].
Unsupervised and lexicon based	[62], [63], [64].

Table 7 show emphasis on supervised learning models that are wildly studied for sentiment analysis in education domain. These models include the Decision Tree (DT), Support Vector Machine (SVM), K Nearest Neighbor (KNN), Naïve Bayes (NB) and Neural Network (NN).

Table 7 Supervised learning models that are wildly studied for sentiment analysis in education domain

Supervised learning models	Papers
DT	[7], [16], [18], [19], [24], [25], [41], [56], [59].
SVM	[2], [7], [8], [9], [12], [15], [16], [17], [19], [20], [21], [24], [25], [38], [41], [48], [51] [52], [56], [57], [59], [60], [64].
KNN	[9], [15], [18], [19], [23], [52], [64].
NB	[9], [12], [15], [16], [17], [20], [21], [22], [23], [24], [25], [26], [34], [38], [41], [51], [52], [56], [57], [58], [59], [64], [65].
NN	[9], [11], [13], [14], [17], [19], [23], [38], [41], [59], [61].

Additionally, as shown in Table 6, lexicon-based learning approaches, also referred to as rule-based sentiment analysis, were frequently used in a number of research and were frequently linked to either supervised learning techniques or unsupervised learning techniques. We observed that the Valence Aware Dictionary and Sentiment Reasoner (VADER) and Sentiwordnet were used far more frequently than TextBlob, MPQA, Sentistrength, and Semantria in Table 8's list of the most commonly used lexicons elaborated among the examined publications.

Table 8: Frequently used lexicons

Lexicon-Based	Papers
VADER	[36], [38], [42], [43], [48].
Sentiwordnet	[46], [56], [57], [65].
Semantria	[45], [58].
Sentistrength	[44].
TextBlob	[38], [49].
MPQA	[20].

RQ3. What are the most common metrics for measuring the effectiveness of sentiment analysis systems?

Systems designed for sentiment analysis were commonly evaluated using metrics based on information retrieval such as precision, F1-score, and recall. Additionally, other research used measures based on statistics to evaluate the precision of systems.

Comparing the number of articles that utilized a certain assessment measure to evaluate the performance of systems with the number of articles that either performed no evaluation or chose not to stress the employed metrics is highly intriguing. Table 9 shows the percentage of articles defined for each of the assessment metrics.

Table 9: Percentage of evaluation metrics applied in the reviewed papers

Evaluation metrics	Information retrieval-based metrics (accuracy, precision, f1-score, and recall)	Kappa	Pearson R-value	N/A
Papers (%)	67%	4%	3%	26%

Table 9 shows that 67% of the publications either simply featured the accuracy or other evaluation metrics such as the precision, recall, and F1-score. Kappa on the other hand was employed in just 4% of the research, while Pearson R-value in 3%, and no assessment metrics were specified in 26% of the research.

RQ4. What are the most popular methods for gathering student feedback?

During the process of reviewing the papers in this study, we found different data source and we divided them into three different categories based on their characteristics. These categories are:

Questionnaires/Survey: This category of dataset was collected by providing questionnaires in order to gather feedbacks from students or conducting survey among teachers and students.
Social media and blogs: This category of dataset comprises od data that are collected through social media platforms like Facebook, Twitter, and blogs
Education/research platforms: In this dataset category, data are extracted through online educational and research platforms such as edX, Coursera, ResearchGate, Kaggle, LinkedIn, etc.

Based on the reviewed paper, just about a three-quarter of the of the papers disclosed the data source while about one-third of the papers did not disclosed information about the source of dataset collected. A tabular representation of these papers and dataset source is shown in Table 10.

Table 10: Dataset sources that have been used by the reviewed papers

S/N	Category of dataset	Papers	Description
1	Questionnaires/ Surveys	[7], [10], [19], [22], [23], [29], [30], [31], [35], [40], [42], [43], [51], [52], [57], [58], [59], [61].	This category of dataset was collected by providing questionnaires in order to gather feedbacks from students or conducting survey among teachers and students.
2	Social media and blogs	[12], [21], [32], [34], [39], [41], [46], [47], [48], [59], [60], [62], [64], [66].	This category of dataset comprises od data that are collected through social media platforms like Facebook, Twitter, and blogs
3	Research platforms/Education	[8], [9], [13], [24], [25], [26], [28], [50], [56],	In this dataset category, data are extracted through online educational and research platforms such as edX, Coursera, ResearchGate, Kaggle, LinkedIn, etc.

We observed that there are areas in students’ feedback sentiment analysis that need more research and development, one of this area from RQ1 is the use of figurative speeches from students’ feedback, such figurative speeches include the use of irony and sarcasm, this area is lacking and in need of further studies. In RQ2, we observed that most of the techniques used are domain specific and do not perform well in multiple domains. Another challenge from RQ2 is an inability to handle complex construct such as abbreviations and words that has multiple meanings. In RQ4, most of the datasets in the reviewed paper are unstructured, therefore to identify the main entities to which the sentiments were directed is not feasible until after applying an entity extraction model, which makes the application of the existing dataset limited.

As authors explore papers from Scopus, Science Direct, EBSCO, Web Science, IEEE Xplore, ACM DL and SpringerLink, it is possible that some relevant papers from other database have been missed. Also, the research team analysis was done based on the selected papers that was reviewed, while other research has been done with respect to techniques and methods as well as technologies and tools that are employed in sentiment analysis.

From our review study, we were able to identify the major student feedback aspects in sentiments analysis, and based on the paper reviewed we observed that highest rate which is 76% are towards teacher while only 8% are towards institutions and the remaining 16% are towards courses and institutions. Furthermore, we identify five (5) techniques that are majorly researched for using sentiment analysis in the area of education, these techniques include the supervised learning, unsupervised learning, lexicon-based, supervised & lexicon-based, and unsupervised and lexicon-based. Five (5) machine learning algorithms were also identifies with the supervised learning approach, these algorithms include, Decision Tree, Support Vector Machine, K-Nearest Neighbor, Naive Bayes and Neural Network. The lexicons that were associated with the lexicon-based approaches from the reviewed papers are VADER, Sentiwordnet, Semantria, Sentistrength, TextBlob and MPQA. Also, we identified the most common metrics for measuring the effectiveness of sentiment analysis systems to be Information retrieval-based evaluation metrics (as Accuracy, Precision, F1-score, and Recall), Kappa and Pearson R-value. We observed that 26% of the papers reviewed did not use any evaluation metrics while a high percentage (67%) of the papers used Information retrieval-based evaluation metrics, while Kappa and Pearson R-value were reviewed by 4% and 3% respectively. Finally, we were able to identify the most popular methods for gathering student feedback to be through questionnaires/survey, social media and blogs, and through education/research platforms.

Based on the challenges and gaps that has been identified in the revied paper we recommend future research on the following aspects.

Dataset size and structure: majority of the papers that were revied in this research used a small size dataset with less than five thousand samples and this affected the results [67], so future research can work on larger dataset in other to make the result more reliable. Also, a structured format feedback dataset is needed via a survey and questionnaire, rather than the unstructured format that is being used.
Emotion Detection: Only few of the articles that were reviewed focused on detecting students’ emotion for sentiments analysis. Thus, we recommend a future work that consider the use of students’ emotion expression as feedback for the purpose of student sentiment analysis.

Ethics approval: Ethics approval is not applicable.

Consent to Participate: All the Authors in this study agreed and freely participated in the preparation of the manuscript.

Consent for publication: All the Authors agreed that this work and data therein should be published in the Journal of Big Data.

Availability of data and materials: The data and materials that supports the findings of this paper are available and can be obtained from the corresponding author upon reasonable request.

Competing interests: I have the consent of all the co-authors to submit this work to your Journal and I can assure you that there is no contesting interest to this publication.

Funding: This research received no internal or external funding

Authors' contributions: The manuscript was written through the contributions of all the authors. Corresponding Author was responsible for the conceptualization of the topic. Article gathering and sorting was carried out by the third author. Manuscript writing and original drafting and formal analysis were carried out by the corresponding author and second author. Reviews and editing of the manuscript were carried out by the third and fourth authors. All the Authors participated in writing of the manuscript and overall research activities.

All the Authors have read and agreed to publish the manuscript.

Acknowledgement: We really want to thank and acknowledge God Almighty for making it possible for us to complete this manuscript.

We sincerely thank the Management and staff of the Federal University, Lokoja for creating a peaceful learning environment devoid of intimidation, harassment, and other forms of crime, I pray and wish that peace should continue to exist in this University.

Authors' information (optional):

1^st Author: Emughedi Oghu Master Student, Department of Computer Science, Federal University, Lokoja, Kogi State, Nigeria.

2^nd Author: Emeka Ogbuju, Department of Computer Science, Federal University, Lokoja, Kogi State, Nigeria.

3rd Author: Taiwo Abiodun Master Student, Department of Computer Science, Federal University, Lokoja, Kogi State, Nigeria.

4^th Author: Francisca Oladipo Department of Computer Science, Federal University, Lokoja, Kogi State, Nigeria.

Plutchik, R. (2001). The Nature of Emotions. American Scientist, 344-350.
Mite, B. K., Delgado, V. C., Solís, A. E., Espinoza, A. H., Ortiz, Z. J., & Varela, T. E. (2018). Sentiment Analysis in Education Domain: A Systematic Literature Review. International Conference on Technologies and Innovation (pp. 285-297). Berlin Germany: Springer International Publishing.
Han, Z., Wu, J., Huang, C., Huang, Q., & Zhao, M. (2020). A review on sentiment discovery and analysis of educational big-data. Wiley Interdiscip. Rev. Data Min. Knowl. Discov.
Zhou, J., & Min, Y. J. (2020). Sentiment analysis in education research: A review of journal publications. Interact. Learn. Environ, 1-13.
Kitchenham, B. A., & Charters, S. (2007). Guidelines for performing systematic literature review in software engineering. Technical report 2(3), Keele University and University of Durham.
Host, M. & Orucevic-Alagic, A. (2013). A systematic review of research on open source software in commercial software product development.
Dake, D. K., & Gyimah, E. (2022). Using Sentiment analysis to evaluate qualitative students' responses. Edu Inf Technol.
Faizi, R., & Fkihi, S. E. (2022). A sentiment analysis based approach for exploring student feedback. In Innovative Technologies and Learning (pp. 52-59).
Mabunda, J., Jadhav, A., & Ajoodha, R. (2021). Sentiment analysis of student textual feedback to improve teaching. International conference on Interdisciplinary Research in Technology and Management. Kolkata, India: IRTM.
Nandakumar, R., Pallavi, M. S., Haeithan, P. P., & Hegde, V. (2022). Sentiment Analysis on Student Feedback using NLP & POS Tagging. International Conference on Edge Computing and Applications (pp. 309-313). IEEE.
Peng, H., Zhang, Z., & Liu, H. (2022). A sentiment analysis method for teaching evaluation texts using attention mechanism combined with CNN-BLSTM Model. Hindawi.
Umair, M., Hakim, A., Hussain, A., & Naseem, S. (2021). Sentiment analysis of students' feedback before and after the COVID-19 pandemic. International Journal of Emerging Technology, 177-182.
Katragadda, S., Ravi, V., Kumar, P., & Lakshmi, G. J. (2020). Performance Analysis on Student Feedback using Machine Learning Algorithms. In Proceedings of the 2020 6th International Conference on Advanced Computing and Communication Systems (ICACCS), (pp. 1161–1163). Coimbatore, India.
Estrada, M. L., Cabada, R. Z., Bustillos, R. O., & Graff, M. (2020). Opinion mining and emotion recognition applied to learning environments. journal of Expert System with Applications, 113265.
Hew, K. F., Hu, X., Qjao, C., & Tang, Y. (2020). What predicts student satisfaction with MOOCs: A gradient boosting trees supervised machine learning and sentiment analysis approach. Computers & Education, 103724.
Giang, N. T., Dien, T. T., & Khoa, T. T. (2020). Sentiment Analysis for University Students’ Feedback. Future of Information and communication conference (pp. 55-66). Berlin, Germany: Springer.
Katragadda, S., Ravi, V., Kumar, P., & Lakshmi, G. J. (2020). Performance Analysis on Student Feedback using Machine Learning Algorithms. In Proceedings of the 2020 6th International Conference on Advanced Computing and Communication Systems (ICACCS), (pp. 1161–1163). Coimbatore, India.
Kavitha, R. (2019). Sentiment Research on Student Feedback to Improve Experiences in Blended Learning Environments. International Journal of Innovation.
Spatiotis, N., Perikos, I., Mporas, I., & Paraskevas, M. (2018). Evaluation of an educational training platform using text mining. In Proceedings of the 10th Hellenic Conference on Artificial Intelligence, (pp. 1-5). Patras, Greece.
Terkik, A., Prud’hommeaux, E., Alm, C. O., & Homan, C. (2016). Analyzing gender bias in student evaluations. In Proceedings of the COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, (pp. 868–876). Osaka, Japan.
Esparza, G. G., de Luna, A., Zezzatti, A. O., Hernandez, A., Ponce, J., Álvarez, M., . . . de Jesus, N. (2017). A sentiment analysis model to analyze students reviews of teacher performance using support vector machines. In International Symposium on Distributed Computing and Artificial Intelligence (pp. 157-164). Berlin/Heidelberg, Germany: Springer.
Van, N. K., Nguyen, V. D., Nguyen, P. X., Truong, T. T., & Nguyen, N. (2018). Uit-vsfc: Vietnamese students’ feedback corpus for sentiment analysis. . In Proceedings of the 2018 10th International Conference on Knowledge and Systems Engineering, (pp. 19-24). HoChi Minh City, Vietnam.
Koufakou, A., Gosselin, J., & Guo, D. (2016). Using data mining to extract knowledge from student evaluation comments in undergraduate courses. In Proceedings of the 2016 International Joint Conference on Neural Networks, (pp. 3138–3142). Vancouver, Canada.
Sultana, J., Sultana, N., Yadav, K., & AlFayez, F. (2018). Prediction of sentiment analysis on educational data based on deep learning approach. In Proceedings of the 2018 21st Saudi Computer Society National Computer Conference, (pp. 1-5). Riyadh, Saudi Arabia.
Ibrahim, Z. M., Bader-El-Den, M., & Cocea, M. (2018). A data mining framework for analyzing students’ feedback of assessment. In Proceedings of the 13th European Conference on Technology Enhanced Learning Doctoral Consortium, (p. 13). Leeds, UK.
Krishnaveni, K., Pai, R. R., & Iyer, V. (2017). Faculty rating system based on student feedbacks using sentimental analysis. In Proceedings of the 2017 International Conference on Advances in Computing, Communications and Informatics, (pp. 1648–1653). Udupi, India.
Ortega, M. P., Mendoza, L. B., Hormaza, J. M., & Soto, S. V. (2020). Accuracy measures of sentiment analysis algorithms for spanish corpus generated in peer assessment. In Proceedings of the 6th International Conference on Engineering & MIS 2020, (pp. 1-7). Larnaka, Cyprus,.
Lwin, H. H., Oo, S., Ye, K. Z., Lin, K. K., & Aung, W. P. (2020). Feedback Analysis in Outcome Base Education Using Machine Learning. In Proceedings of the 2020 17th International Conference on Electrical Engineering/Electronics, Computer, Telecommunications and Information Technology (pp. 767–770). Phuket, Thailand: .
Jimenez, H. G., Finamore, C. A., & Simoes, G. (2021). Sentiment analysis of student surveys: A case study on assessing the impact of theCOVID-19pandemic on highereducation teaching. Proceedings of the 14th International Conference on Education Data Mining (pp. 353-359). EMD.
Cunningham-Nelson, S., Baktashmotlagh, M., & Boles, W. (2019). Visualizing Student Opinion Through Text Analysis. IEEE Transactions on Education, 305-311.
Chantamuang, O., Polpinij, J., Vorakitphan, V., & Luaphol, B. (2022). Sentence-Level sentiment analysis for student feedback relevant to teaching process assessment . In proceedings of Multi-disciplinary Trends in Artificial Intelligence, (pp. 156-168).
Sivakumar, M., & Reddy, U. S. (2017). Aspect based sentiment analysis of students opinion using machine learning techniques. Proceedings of the 2017 International Conference on Inventive Computing and Informatics (ICICI) (pp. 726–731). Coimbatore, India: IEEE.
Nitin, G. I., Swapna, G., & Shankararaman, V. (2015). Analyzing educational comments for topics and sentiments: A text analytics approach. Proceedings of the 2015 IEEE Frontiers in Education Conference (FIE) (pp. 21-24). texas, USA.: IEEE.
Barrón-Estrada, M. L., Zatarain-Cabada, R., Oramas-Bustillos, R., & González-Hernández, F. (2017). Sentiment analysis in an affective intelligent tutoring system. In Proceedings of the 2017 IEEE 17th international conference on advanced learning technologies (pp. 394–397). Timisoara, Romania: IEEE.
Chauhan, G. S., Agrawal, P., & Kum, Y. (2019). Aspect-Based Sentiment Analysis of Students’ Feedback to Improve Teaching–Learning Process. Information and Communication Technology for Intelligent Systems, 259–266.
Martinis, M. C., Zucco, C., & Cannataro, M. (2022). An Italian lexicon-based sentiment analysis approach for medical applications. In proceedings of the 13th ACM International Conference on Bioinformatics (pp. 1-4). New York, NY, United State: Association for Computing Machinery.
Rajput, Q., Haider, S., & Ghani, S. (2016). Lexicon-Based Sentiment Analysis of Teachers' Evaluation. Appl. Comput. Intell. Soft Comput.
Cobos, R., Jurado, F., & Blázquez-Herranz, A. (2019). A Content Analysis System that supports Sentiment Analysis for Subjectivity and Polarity detection in Online Courses. IEEE, 177–187.
Pong-Inwong, C., & Songpan, W. (2019). Sentiment analysis in teaching evaluations using sentiment phrase pattern matching (SPPM) based on association mining. International Journal of Machine Learning , 2177–2186.
Herath, M., Chamindu, K., Maduwantha, H., & Ranathunga, S. (2022). Dataset and baseline for authomatic student feedback analysis. Proceedings of the Thirteenth Language Resources and Evaluation Conference (pp. 2042-2049). Marseille, France: European Language Resources Association.
Nguyen, V. D., Van, N. K., & Nguyen, N. (2018). Variants of Long Short-Term Memory for Sentiment Analysis on Vietnamese. . In Proceedings of the 2018 10th International Conference on Knowledge and Systems Engineering, (pp. 306–311.). Ho Chi Minh City, Vietnam,.
Newman, H., & Joyner, D. (2018). Sentiment analysis of student evaluations of teaching. In International Conference on Artificial Intelligence in Education (pp. 246–250). Berlin, Germany: Springer.
Raj, A. G., Ketsuriyonk, K., Patel, J. M., & Halverson, R. (2017). What Do Students Feel about Learning Programming Using Both English and Their Native Language. In Proceedings of the 2017 International Conference on Learning and Teaching in Computing and Engineering (LaTICE), (pp. 1-8). Hong Kong, China,.
Jiranantanagorn, P., & Shen, H. (2016). Sentiment analysis and visualisation in a backchannel system. In Proceedings of the 28th Australian Conference on Computer-Human Interaction, (pp. 353–357). Launceston, Australia.
Santos, C. L., Rita, P., & Guerreiro, J. (2018). Improving international attractiveness of higher education institutions based on text mining and sentiment analysis. Int. J. Educ. Manag, 431–447.
Iram, A. (2018). Sentiment Analysis of Student’s Facebook Posts. In International Conference on Intelligent Technologies and Applications (pp. 86-97). Berlin/Heidelberg, Germany: Springer.
Liu, Z., Yang, C., Peng, X., Sun, J., & Liu, S. (2017). Joint exploration of negative academic emotion and topics in student-generated online course comments. In Proceedings of the 2017 International Conference of Educational Innovation through Technology, (pp. 89-93). Osaka, Japan.
Elia, G., Solazzo, G., Lorenzo, G., & Passiante, G. (2019). Assessing learners’ satisfaction in collaborative online courses through a big data approach. Computers in Human Behavior, 589–599.
Pyasi, S., Gottipati, S., & Shankararaman, V. (2018). An Analytics Tool for Gaining Insights from Student Feedback Comments. In Proceedings of the 2018 IEEE Frontiers in Education Conference (FIE) (pp. 1-9). San Jose, CA, USA: IEEE.
Lubis, F. F., Rosmansyah, Y., & Supangkat, S. H. (2016). Experience in learners review to determine attribute relation for course completion. In Proceedings of the 2016 International Conference on ICT For Smart Society, (pp. 32-36). Surabaya, Indonesia.
Colace, F., Casaburi, L., De Santos, M., & Greco, L. (2015). Sentiment detection in social networks and in collaborative learning environments. Comput. Hum. Behav, 1061–1067.
Almosawi, M. M., & Mahmood, S. A. (2022). Lexicon-Based approach for sentiment analysis to student feedback. Webology, 6971-6989.
Spatiotis, N., Perikos, I., Mporas, I., & Paraskevas. (2020). Sentiment Analysis of Teachers Using Social Information in Educational Platform Environments. International Journal of Artificial Inteligence.
Sangeetha, K., & Prabha, D. (2020). Sentiment analysis of student feedback using multi-head attention fusion model of word and context embedding for LSTM. Journal of Ambient Intelligence and Humanized Computing , 4117–4126.
Soe, N., & Soe, P. T. (2019). Domain Oriented Aspect Detection for Student Feedback System. In Proceedings of the 2019 International Conference on Advanced Information Technologies (ICAIT), (p. 90-95). Yangon, Myanmar.
Korkmaz, M. (2018). Sentiment analysis on university satisfaction in social media. In Proceedings of the 2018 Electric Electronics, Computer Science, Biomedical Engineerings’ Meeting (EBBT), (pp. 1-4). Istanbul, Turkey.
Mandal, L., Das, R., Bhattacharya, S., & Basu, P. N. (2017). Intellimote: A hybrid classifier for classifying learners’ emotion in a distributed e-learning environment. Turkish Journal of Electrical Engineering and Computer Sciences, 2084–2095.
Muollo, K., Basavaraj, P., & Garibay, I. (2018). Understanding Students’ Online Reviews to Improve College Experience and Graduation Rates of STEM Programs at the Largest Post-Secondary Institution: A Learner-Centered Study. In Proceedings of the 2018 IEEE Frontiers in Education Conference (FIE) (pp. 1-7). San Jose, CA, USA: IEEE.
Sahu, G. T., & Dhyani, S. (2019). Dynamic feature based computational model of sentiment analysis to improve teaching learning system. Int. J. Emerg. Technol, 17-23.
Nasim, Z., Rajput, Q., & Haider, S. (2017). Sentiment analysis of student feedback using machine learning and lexicon based approaches. In Proceedings of the 2017 international conference on research and innovation in information systems, (pp. 1-6). Langkawi, Malaysia.
Sindhu, I., Daudpota, S. M., Badar, K., & Bakhtyar, M. (2019). Aspect-based opinion mining on student’s feedback for faculty teaching performance evaluation. IEEE, 108729–108741.
Srinivas, S., & Rajendran, S. (2019). Topic-based knowledge mining of online student reviews for strategic planning in universities. Computers & Industrial Engineering, 974–984.
Nguyen, P. X., Hong, T. T., Van, N. K., & Nguyen, N. L. (2018). Deep learning versus traditional classifiers on Vietnamese students’ feedback corpus. Proceedings of the 2018 5th NAFOSTED Conference on Information and Computer Science (NICS), (pp. 75-80). Ho Chi Minh City, Vietnam.
Abed, A. A., & El-Halees, A. M. (2017). Detecting subjectivity in staff perfomance appraisals by using text mining: Teachers appraisals of palestinian government case study. In Proceedings of the 2017 Palestinian International Conference on Information and Communication Technology, (pp. 120-125). Gaza, Palestine.
Balachandran, L., & Kirupananda, A. (2017). Online reviews evaluation system for higher education institution: An aspect based sentiment analysis tool. In Proceedings of the 2017 11th International Conference on Software, Knowledge, Information Management and Applications (SKIMA), (pp. 1-7). Malabe, Sri Lanka.
Nimala, K., & Jebakumar, R. (2019). Sentiment topic emotion model on students feedback for educational benefits and practices. Behav. Inf. Technol, 311-319.
Kagklis, V., Karatrantou, A., Tantoula, M., Panagiotakopoulos, C. T., & Verykios, V. S. (2015). A learning analytics methodology for detecting sentiment in student fora: A case study in Distance Education. Eur. J. Open Distance E-Learn, 74-94.

No competing interests reported.

Download PDF

Editorial decision: Major revision
08 Mar, 2023
Reviews received at journal
12 Feb, 2023
Reviewers agreed at journal
01 Feb, 2023
Reviewers agreed at journal
01 Feb, 2023
Reviewers invited by journal
01 Feb, 2023
Editor assigned by journal
25 Jan, 2023
Submission checks completed at journal
12 Jan, 2023
First submitted to journal
11 Jan, 2023

You are reading this latest preprint version

A Review of Sentiment Analysis Approaches for Quality Assurance in Teaching and Learning

Status:

Version 1

Abstract

1. Introduction

2. Background And Related Work

3. Research Method

Eligibility criteria

4. Results

5. Identified Gaps And Challenges

6. Limitation Of The Review

7. Conclusion

8. Further Work

Declarations

References

Additional Declarations

Status:

Version 1