A large amount of user-generated content on social media has led to the pursuit of quickly and accurately mining through data and gathering useful insights. Text sentiment analysis has become a necessary tool in classifying user opinions within Web generated content. Due to the various ways, opinions can be conveyed, performing text sentiment analysis in specific domains becomes a difficult task. With an even greater degree of difficulty added when slang or colloquialisms are used. There is a great deal of research into investigating various classifiers in a traditional natural language processing setting each with their own merits and demerits. In this paper, we present a slang-based dictionary classifier with the objective of determining the sentiment of Instagram comments within the context of fashion, or more specifically sports shoes, and compare it with the performance of other classifiers such as a Naive Bayes, J48, lexicon and random forest. The dataset used for the benchmark was created from popular fashion Instagram accounts. Overall, the random forest classifier yields the best results with an accuracy of 88%, precision of 84% and a recall of 88%.
CITATION STYLE
Aly, E. S., & van der Haar, D. T. (2020). Slang-Based Text Sentiment Analysis in Instagram. In Advances in Intelligent Systems and Computing (Vol. 1027, pp. 321–329). Springer. https://doi.org/10.1007/978-981-32-9343-4_25
Mendeley helps you to discover research relevant for your work.