A sentiment analysis of Twitter posts about news

Abstract

Sentiment analysis of Twitter posts about news sets out to computationally analyse the sentiment of tweets about news. It attempts to find novel ways of extracting tweets about news from Twitter, and it examines whether context plays a role in the determination of the sentiment of tweets about news. Sentiment analysis of tweets about news is a study to do a three-classed (positive, negative, neutral) classification. It does experiments with different operations, feature selections, instance representations, and learning algorithms and recommends the combination that gives improved performance. I believe this research question makes a good departure from the existing sentiment analysis studies.

The thesis is organized as follows. There are 4 more chapters. In chapter 2, methodology and approaches will be presented. Review of literature and methodological approaches will be presented. Under review of literature, general sentiment analysis, difficulty of sentiment analysis, feature engineering, classification, sentiment analysis techniques and sentiment analysis of Twitter messages in particular will be discussed. Under methodological approaches, the literature is interpreted, evaluation is briefly discussed and best methods and approaches are identified. In chapter 3, data collection and preprocessing will be discussed. Under this chapter Twitter APIs, test data, training data, and Twitter posts about news will be discussed. Chapter 4, discusses experiment and results. Here, keyword-based classification, supervised approaches, and preprocessing of data are presented. Under supervised approaches, results for several algorithms using different representation formats are presented. Finally, in chapter 5, analysis and conclusions are provided. Factors affecting performance, evaluations, challenges and conclusions are provided in this chapter.