Abstract
Among the various challenges regarding distance education is the necessity of reducing the student dropout rate. In this sense, the present research aimed to contribute to the design of a lexical database focused on emotions and opinions that can be incorporated into a predictive evasion software. For the database design, we used the Scup tool to collect 150 tweets containing distance education students’ opinions and analyzed them in the light of Martin and White’s Appraisal Framework, along with five resources related to the sentiment Analysis field, which were taken from Liu’s work. In addition, we used the Aulete dictionary to describe the lexical units found in our corpus to better fit them into the analysis categories. Results showed 220 opinion tokens, which were identified and labeled according to their polarity. Moreover, these tokens were included in the domains attitude (judgment and appreciation) and graduation (sharp and strong) from the linguistic framework used. The results also indicated the necessity of another resource to help identify the use of figurative language, slangs, and extralinguistic elements, such as GIFS and emojis.
Highlights
When we think about the future of education, we assuredly think about technology and its innovations
The first domain involves the linguistic description of the phenomena found in our corpus, while the second one is preoccupied with representing the linguistic objects present in the corpus formally; the third domain comprises the encoding of the linguistic information generated in the two first domains, as well as the design of the software
After a careful analysis of our corpus, we found that 84% of the opinions examined were negative, whilst 16% were positive
Summary
When we think about the future of education, we assuredly think about technology and its innovations. This hiatus is purely due to the lack of communication exchange between students and teachers, which might trigger the former to feel abandoned by the latter and isolated from the group and reduce the effectiveness of the class content In this sense, the previously mentioned applications, especially the semantic and artificial intelligence ones, present relevant resources to improve the Distance Education scenario. This research is part of a larger project - MAS-EaD - whose main objective is to design a sentiment analyzer capable of detecting students' sentiments and emotions on a university virtual learning and teaching platform. Thereby, bringing an informal corpus for the analysis in order to seek speech naturalness and new word classes in a different context from which the other corpora were obtained
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
