Real-time Visualisation Of Crowds In Metropolitan Areas
Created a web application using MEAN stack where geo-coded tweets were used for real-time visualisation of areas where there
was a crowd. The website would ask for the name of a city, it would use the twitter API to fetch geo-coded tweets,
we would classify areas with a high density of tweets as a possible crowd, run sentiment analysis on the tweets and identified the topic of interest and the prevailing sentiments in the crowd.
This was built as an aid to the local police to quickly identify areas with violent or angry crowds and take the necessary measures.