Text mining and determinants of sentiments : Twitter social media usage by traditional media houses in Uganda

Abstract

Unstructured data generated from sources such as social media and traditional text documents are increasing and form a larger proportion of un-analysed data, especially in developing countries. In this study, data received from the major print and non-print media houses in Uganda through the Twitter platform was analysed to generate non-trivial knowledge by using text mining analytics. The study aimed at identifying the determinants of derived sentiments in Twitter messaging and evaluate the performance of the model for the determinants of sentiments using topic modelling and a logistic regression model. The results show that sentiments generated from tweets derived from the main print media houses (Daily Monitor and New Vision) were positively correlated (p < 0:05), so were the sentiments from the non-print media (NBS TV and NTV) for the study period. Most Sentiments on security, politics and economics were found to be negative while those on sports were positive. Furthermore, the tweet sentiment statistical logistic model revealed that negative sentiments were determined by the retweeted tweets (OR = 0:889), source of the tweets (web OR < 1), and topics on security (OR = 0:53). Moreover, the positive sentiments were determined by the topic of discussion (sports OR = 8:93 and other unclassified topics OR = 2:3), type of media house (p < 0:05). In conclusion, negative sentiments are drivers of active participation on social media hence the need for constant surveillance. It’s also advisable that for one to get a rich context of the social media content, they should follow one media house per Newspaper (print media) and Television media (non-print media). I recommend an extension on the sentiment model to be based on the concept of big data analytics.