Using Text Mining and Clustering Techniques on Tweets to Discover Trending Topics in Dubai

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
The British University in Dubai (BUiD)
Twitter micro-blogging website is a hot and emerging area of research recently, users on Twitter post millions of tweets every day from all over the world, it is very difficult and challenging task to keep track of messages and filter them based on topical interest. This study uses text mining and clustering techniques to partition Dubai tweets into clusters of a same topical interest. Tweets corpus of Dubai were collected, they were presented through the bag of words model using TF-IDF weighting scheme, after that the output of text transformation was introduced to k-means clustering algorithm with cosine similarity measure. Findings indicate that heuristic evaluation techniques are not so helpful in this domain; also, the model has generated interesting clusters about trending topics and events in Dubai. In the end, an experiment was conducted over datasets collected from different timeframes to see what are the constant hot topics discussed in Twitter about Dubai. All of the findings in this report have been empirically analysed through real-word tweets dataset.
text mining, clustering techniques, tweets, Dubai, United Arab Emirates (UAE), twitter micro-blogging