News: Mediata wins the first prize at IBM Watson hackathon
Mediata’s India based development team, managed locally by Bharath, took some time out of their work on Mediata’s location profiling platform totals part in the IBM hackathon.
Competing with over 80 mix-teams of developers, UX/UI designers, entrepreneurs and creatives, Mediata’s partners in India Bharath, Piyush, Vivek and Sanjitha were challenged to use IBM Watson super computer which is specialized in natural language processing in solving a business problem to create a cognitive computing app.
Named as ‘Sentipede’, Mediata’s team built a product to act as a tool that helps brands like flipkart.com or Vodafone.com to automate the identification of their customers’ who are dissatisfied with the service and have posted about it on social media.
They did this by mining tweets and doing a sentiment analysis (as negative or positive sentiments) via IBM Watson APIs and then figuring out the geo location and the IP from where the tweet is originating to add a customized message on the service website for those IPs.
This message (like an ad banner) could act as an apology/acknowledgement and promise for a better service specifically to those specific negative tweeting customers (or their location). This can also be used to find the negative sentiments of competitors and to strategize on acquiring them as potential customers for any brand.
“Our Profile Engine breaks down dozens of unstructured data sources to create a cohesive profile for any location in ANZ. We use this knowledge to help out advertisers pre-target their campaigns across multiple media types. Seeing the team apply the knowledge from this tool to win 1st prize at such a reputable hackathon is a great source of pride for the whole company.” – Bharath
Mediata’s team got very positive feedback from judges and a room full of curious audience in the final pitch and overall was a fun experience.
At Mediata, part of our everyday job is mining lots of data and understanding and modelling them for relevant advertising targeting. The team were well prepared as a habit to mine data when started with the Hackathon.
The tool we built helps us in enhancing the social data layer of our profile engine by understanding any keyword on social media and finding the sentiment of the person who posted those keywords on social media.