Sentiment analysis of tweets in comparison to a company’s financial performance

Detta är en Kandidat-uppsats från KTH/Skolan för elektroteknik och datavetenskap (EECS); KTH/Skolan för elektroteknik och datavetenskap (EECS)

Sammanfattning: This study analyzes peoples reactions in social media to the release of a company’s quarterly report. Sentiment analysis was performed on tweets about a company both from a short-and long-term perspective. On the long-term perspective, a two year period of sentiment was studied in relation to its quarter’s percental change of net income and revenues. Three companies was investigated in this process. On the short-term perspective, a linear regression was conducted based the quarter’s percental change of net income and revenues and on sentiment score, ranging from release day to 2 days after the release of a company’s quarterly report, in total five companies were studied. The result inferred that there was no correlation between the company’s net income and the reaction on Twitter on both long and short-term. Our conclusion is that the attitude towards a company is influenced by other factors than net income and revenues. The study also indicates that sentiments found in Twitter messages regarding a company name are related to a positive or negative expectation of the company. Findings suggest future studies to focus on companies greatly dependent on one product and analyze sentiment regarding that product instead, as the product is likely to impact financial results.

  HÄR KAN DU HÄMTA UPPSATSEN I FULLTEXT. (följ länken till nästa sida)