Client uses customer media data for targeted advertising and provides bleeding-edge TV solutions. It involves audience segmentation, micro-targeting and in-depth analytics. However, the audience data is big, fast moving, comes from diverse sources and is noisy.


The Machine Learning Hub curated human-like errors by de-noising TV/Cable programming metadata, performing string matching and clustering, and applying this to other domains as well.


  • Media

Tools Used

  • Data Science Experience (DSX)

Data Science Techniques

  • Unsupervised learning & control theory


The result was an increase in data quality, leading to better advertisement targeting. Customers now get more relevant advertisements, based upon their behavioral patterns and preferences.

Need help with a similar case?

Contact us to find out how we can help!