2026-07-27T01:32:58Z https://meral.edu.mm/oai

oai:meral.edu.mm:recid/4812 2021-12-13T01:05:04Z 1582963302567:1597824273898 user-ucsy

Sentiment Analysis for Twitter Stream Data by Combining Lexicon and Machine Learning Approaches Chan, Wint Nyein Thein, Thandar Nowadays, Twitter Sentiment analysis hasbecome popular as it helps the organization todetermine marketing strategy by providing publicopinions. Efficient techniques to collect a large amountof Twitter stream data and extracting sentimentinformation from collected raw data are essentialdemand. Traditional sentiment classificationtechniques do not perform well in Social Data.Acquiring effective training data is a challengealthough learning based approaches are good forSocial Data Sentiment Classification. Manual Labelingfor training data is time and labor consuming. In thispaper, Sentiment Analysis System for Twitter data isproposed with five modules: Data Collection,Preprocessing, Class Labeling, Classification ModelDevelopment and Sentiment Classification. TheSentiment Classification is implemented by combininglexicon and Supervised learning-based approaches. Inthis system, lexicon-based classifier is applied to labelthe class and suitable learning-based classifier ischosen for classification. Emoticon and slang wordsare considered for classification. To select suitableclassifier, three different classification algorithms areevaluated. The performance evaluation shows thatNaïve Bayes classifier is better and the proposedsystem can achieved the promising accuracy. 2017-02-17 http://hdl.handle.net/20.500.12678/0000004812 https://meral.edu.mm/records/4812