Project SMART (Social Media Analytics in Real Time)     


Wanita Sherchan photo

Project SMART (Social Media Analytics in Real Time) - overview

Social Media Analytics in Real Time, also known as Project SMART, is an IBM Research solution that analyses and visualises the Twitter stream in real time for any topic of interest. Roughly 9,100 tweets are analysed every second, of which 36% contain photo images.

Monitoring and analysis of streaming data, such as social media, sensors, and news feeds, has become increasingly important to our clients. This asset helps clients analyse what's happening on social media in regard to their target topic, which, for example, could be a product, organisation or an event etc. IBM research developed SMART, which provides a web architecture for real-time analytics of streaming Twitter data. On entering a search term on the Project SMART page, the user is presented with a set of visualisations on: the number and rate of tweets, top words with calculated sentiment, word cloud representing the frequency of words, geolocation prediction of tweets, clustering of tweets into topics and the tweets themselves. Most recently we enhanced SMART with an image analytics widget, which applies text processing analytics and semantic categorisation to twitter images that are most relevant to the given topic.

Project SMART has been used to feed social media data into the CrowdTracker for Australian Open in 2015. It has also been successfully deployed for tracking social media data during Melbourne Spring Fashion Week 2014, and is currently deployed in an engagement with an Australian Insurance Company.