Spatio-Temporal Visualization Model for Movie Success Prediction Based on Tweets

Amali Wijekoon, Thanuja Chandani Sandanayake, K. D. A. A. Jayawardena, A. L. Y. Buddhini, U. K. D. G. S. Ariyawansha · 2017

Visualization of spatio-temporal data has always been a tricky task due to the multi dimensionality. Developing a powerful spatio-temporal visualization requires, use of several visualization techniques and advanced programming algorithms. But successful implementation of a good visualization can help people to come to quick and accurate decisions. The best example for the practical usage of such visualizations would be the movie industry. Nowadays, social media networks play a major role in fostering communication across geographical boundaries and have been able to capture the general public in an emotional level by providing the facility of openly sharing user opinions. Businesses have taken this concept as an advantage to grab the consumer insights and turned it into an opportunity to improve themselves. So as the movie industry of Hollywood and Bollywood. While Digital Marketing has become the easiest and the most cost-effective way of positioning the product in the targeted segment, businesses are now using social media platforms like Twitter as a valuable source to collect real insights of the aggregate sentiment generated by the movie watching community. With compared to other ubiquitous platforms like Facebook, Twitter is the most well-known platform for opinion sharing. This brings up a requirement to validate these insights that are analyzed from the twitter platform. In addition, a platform like Twitter provides geographical as well as timely information that can be utilized by directors/producers and even actors to understand how their popularity or lack of popularity was propagated over time and space. This research has developed a spatio-temporal model with regard to the association with time, location and several other dimensions which can be extracted from the original data.

Read the paper · More papers on PaperTik