Document Type : Original Article

Authors

1 Database Department, Computer Science Institute, Sulaimani Polytechnic University, Kurdistan Region – Iraq

2 Information Technology Department, Chamchamal Technical Institute, Sulaimani Polytechnic University, Kurdistan Region – Iraq

3 Information Technology Department, Computer Science Institute, Sulaimani Polytechnic University

Abstract

Widespread internet use and the web have brought about new ways of expressing individual sentiments. A sentiment is defined as an individual's view in which feelings, attitudes, and thoughts can be represented. When it comes to analysing and extracting Sentiment analysis and opinion mining are two of the most prominent disciplines of research. They derive insights using text data through numerous sources like Facebook and Twitter. Sentiment analysis frequently elicits information on how people feel about various events, brands, products, or businesses. Researchers collect and improvise replies from the general public to conduct evaluations. This paper looks into sentiment analysis for classifying Twitter subscriber tweets. This approach can help analysing the information gathered and stored in positive, neutral and negative opinions. This information is first pre-processed before creating feature vectors. On the basis of machine learning, classification methods were used. The study's algorithms are used Maximum Entropy, Naive Bayes and Support Vector Machine; they are used to categorize documents as positive or negative. The dataset for this paper are obtained from Twitter and includes subscribed tweets by using the API. Following pre-processing, machine learning methods are used to determine whether the tweets are positive or negative.

Keywords

Main Subjects