Please use this identifier to cite or link to this item:
https://doi.org/10.21256/zhaw-21754| Publication type: | Bachelor thesis |
| Title: | Analysing student comments on RateMyProfessors.com using NLP techniques |
| Authors: | Gioli, Giacomo |
| Advisors / Reviewers: | Fazlija, Bledar |
| DOI: | 10.21256/zhaw-21754 |
| Extent: | 58 |
| Issue Date: | 2020 |
| Publisher / Ed. Institution: | ZHAW Zürcher Hochschule für Angewandte Wissenschaften |
| Publisher / Ed. Institution: | Winterthur |
| Language: | English |
| Subject (DDC): | 378: Higher education 410.285: Computational linguistics |
| Abstract: | The assessment of teaching methods and faculty performance is an important step enabling educational institutions to continuously improve their teaching methods and study offers. Typically, schools conduct internal surveys to assess their performance. In most cases, however, the results of these surveys are not disclosed to the public. Therefore, several online platforms have emerged, which allow students to evaluate their teachers publicly. The most popular online evaluation platform, RateMyProfessors.com, currently features over 15 million evaluations covering more than 1.8 million teachers. So, how can schools use this large amount of publicly available data to generate useful insights? In order to answer this research question, a dataset containing 1,637,435 evaluations for 134,375 teachers from 605 schools selected with a random approach using web scraping techniques was built. Intermediate questions were defined in order to answer the research question, such as whether it is possible to use computational techniques to distinguish good from bad teachers based on the language used by the students. The individual questions were elaborated and answered using theoretical knowledge and statistical models. Using natural language processing and machine learning techniques it was demonstrated that it is possible to distinguish positive evaluations from negative evaluations, easy subjects from difficult subjects as well as good teachers from bad teachers with accuracies of over 90%. Furthermore, thanks to the correlations discovered between the quality of teaching as perceived by students, the level of difficulty as perceived by students and the helpfulness of the teacher, it was possible to predict the quality of teaching and the level of difficulty based on the students language. Finally, it was demonstrated that, using statistical models, it is possible to identify topics concerning the faculty performance and teaching methods in evaluations of online courses. Although random approaches to data collection have been chosen to allow the results to be generalized, this cannot be considered universally valid, as the platform from which the data was extracted offers the possibility to evaluate only institutes in the United States, Canada and the United Kingdom. It is therefore necessary to consider possible differences in the way teachers in other cultures are evaluated. In conclusion, natural language processing and machine learning techniques can be applied for the analysis of online evaluations. Schools can therefore use these techniques to generate useful information about their teachers and their teachers’ performance based on online evaluations. This approach, however, should not be looked at by schools as an alternative to the typical evaluation activity, but as an extension, allowing them to analyze aspects not normally considered in typical school evaluations. |
| URI: | https://digitalcollection.zhaw.ch/handle/11475/21754 |
| License (according to publishing contract): | CC BY 4.0: Attribution 4.0 International |
| Departement: | School of Management and Law |
| Appears in collections: | BSc Wirtschaftsinformatik |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| gioligia_thesis.pdf | 7.74 MB | Adobe PDF |  View/Open |
Show full item record
Gioli, G. (2020). Analysing student comments on RateMyProfessors.com using NLP techniques [Bachelor’s thesis, ZHAW Zürcher Hochschule für Angewandte Wissenschaften]. https://doi.org/10.21256/zhaw-21754
Gioli, G. (2020) Analysing student comments on RateMyProfessors.com using NLP techniques. Bachelor’s thesis. ZHAW Zürcher Hochschule für Angewandte Wissenschaften. Available at: https://doi.org/10.21256/zhaw-21754.
G. Gioli, “Analysing student comments on RateMyProfessors.com using NLP techniques,” Bachelor’s thesis, ZHAW Zürcher Hochschule für Angewandte Wissenschaften, Winterthur, 2020. doi: 10.21256/zhaw-21754.
GIOLI, Giacomo, 2020. Analysing student comments on RateMyProfessors.com using NLP techniques. Bachelor’s thesis. Winterthur: ZHAW Zürcher Hochschule für Angewandte Wissenschaften
Gioli, Giacomo. 2020. “Analysing Student Comments on RateMyProfessors.com Using NLP Techniques.” Bachelor’s thesis, Winterthur: ZHAW Zürcher Hochschule für Angewandte Wissenschaften. https://doi.org/10.21256/zhaw-21754.
Gioli, Giacomo. Analysing Student Comments on RateMyProfessors.com Using NLP Techniques. ZHAW Zürcher Hochschule für Angewandte Wissenschaften, 2020, https://doi.org/10.21256/zhaw-21754.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.