TEXT ANALYTICS OF OPINION-POLL ON ADOPTION OF DIGITAL COLLABORATIVE TOOLS FOR ACADEMIC PLANNING USING VADER-BASED LEXICON SENTIMENT ANALYSIS

Abstract

The fast growing community of digital collaborative users across the globe continued to witness breaking of new frontiers in hitherto industries where deployment of traditional methods of computing had continued to hold sway. Notwithstanding the widespread deployment computing tools in educational institutions in Nigeria, the use of online collaborative tools is limited and seldom a commonplace in tertiary educational institutions for academic planning. This study therefore aims at extracting emotions from opinions expressed by stakeholders in the academic research industry regarding the utilitarian possibilities of collaborative tools for academic planning purposes through text mining. A VADER-based approach to Sentiment Analysis is modeled in the opinion mining study of the natural language processing use case. Assigning negative, positive, neutral and compound values to the uni-gram and bi-gram tokenized dictionary-of-known-words, experimental result shows a -0.10 mean sentiment negative score constituting a 17.27% clusters of respondents not favorably disposed to the idea while a 22.7% cluster of highly convinced respondents expressed positive sentiments about the use of collaborative tools with a mean sentiment score of 0.49. A 60.01% cluster of average respondents who expressed neutral sentiments actually tilts towards a positive emotion with a 0.39 mean score.