Answers are grounded in this paper and cite their sources.
Based on current data, there has been an increase in social media users, which shows that more and more people are using social media as a place to express themselves and their emotions. This will generate thousands of tweets within a day. The tweet data is processed so that it is useful for stakeholders who need it to help them make a decision. Because sentence structures on social media are often irregular, pre-processing is necessary to make tweet sentences normal. Stemming and Stopwords are pre-processing techniques that are widely used in sentiment analysis. In previous studies, there were indications that its use did not have a significant effect on accuracy. In this study, the authors divide it into four models: using stemming and stopwords and without using stemming and stopwords. Data using stemming gets the best results with an f1-score of 65%. These results indicate an increase in performance in the use of stemming and stopwords using Multi-class Naive Bayes
No takes yet. Share an insight, caveat, or question.
Jefriyanto et al. (2023) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: