Mikro Bloglardaki Finans Toplulukları için Kullanıcı Ağırlıklandırılmış Duygu Analizi Yöntemi

Nowadays, sentiment analysis is a popular research area in computer science. which aims to determine a person’s or a group’s mood, behaviour and opinion about any topic with using textual documents. With the proliferation of social micro-blogging sites, digital opinion text data is increased. Thus, many sentiment analysis researches are performed on these public data in different sociological fields, such as finance, economy and politics. In this paper, a novel sentiment analysis method is proposed on micro-blogging sites which uses new user metrics. Proposed method is used to measure financial community’s sentiment polarity on micro-blogging sites. In addition to that we analyze the correlation between the mood of financial community and the behavior of the Borsa Istanbul 100 index weekly by Pearson correlation coefficient method. Our test results show that this novel sentiment analysis method improves to the accuracy of linear relationship between the behavior of the stock market and the sentiment polarity of the financial community.