Web Media Mining

Computer Science PhD Course

Course Description

Which videos should YouTube recommend you? Can we answer a question with an image or a video? How to summarize the images or videos published under an hashtag?

This course addresses research problems that aim to understand and mine the richness of Web information to solve important real-world challenges.

We first study computational representations of Web and media data. These representations must be suitable for efficient indexing and to be processed by analysis algorithms.

Extracting and modeling information from Web and media data is then the second part of this course. We will understand how to extract semantic information from Web data and how to model the distribution of semantic information on data.

A major goal of this course is to learn how to deploy information processing algorithms to address large-scale datasets.


  • Understand the state-of-the-art challenges in IR.
  • Deploy massive data processing algorithms.
  • Learn how to analyse and model visual and text information.
  • Design experimental setups and protocols.


Lab work (50%) + Project (50%)


Joao Magalhaes (jmag@xfct.unlx.pt - remove the 'x's to mail us)