Social media services such as Facebook, Flickr and Google+ allow a single user to share multiple images, videos and comments. This generates a large volume of data, presenting a great challenging for information retrieval tasks. Unlike conventional data as numbers and short strings, for images and videos make no sense to compare them in their original domain. This kind of data is known as complex data. Similarity search has been proven to be the state-of-the-art to compare complex data. However, the concept of similarity itself becomes unclear when objects under comparison are composed of parts that can be compared by similarity as well, for example, a user's photo album. This project aims at developing techniques to compare users on social network by considering their content as complex objects. The internship project will be hosted at the Carnegie Mellon University and supervised by Prof. Christos Faloutsos.
News published in Agência FAPESP Newsletter about the scholarship: