Vlado Menkovski bio photo

Vlado Menkovski

Assistant Professor at Eindhoven University of Technology.

Email Google Scholar
  • Web profiling – Based on user behavior and content of websites, are we able to determine demographic profile traits of the user that is browsing the website. For example, can we distil gender or age from a visitor’s behavior?
  • Cross-context attribution models – Churn models, retention models, buyer intent all exist in a specific well-defined area such as clickstreams from one domain, CRM data or specific conversations. Combining those contexts is far more powerful and can be done using attribution models to parameterize the importance of each of the separate contexts. The assignment is to find a generic framework to attribute clickstream data with textual data in such a way that it attributes to a single metric (for example churn prediction).
  • Mobile profiling – Using an SDK shipped with mobile apps, can we learn about the user that is using the mobile phone. A US-based company called PushSpring does this by using a large database mapping apps to demographic traits. The project is to distil a similar mapping but then using machine learning on the user’s behavior, installed apps and profile information from other sources than mobile.