---

Approximate ranking in large graphs

Data objects in many domains are not independent entities. Modeling, searching and analysis of graphs has proved to be a suitable paradigm for such domains. In this project I am investigating bounds approximations of importance computation for large-scale graphs. Experiments, using the developed approximation bounds, will be performed on biological and social networks.

Gene function prediction

A gene network models the interactions between genes in the cell. Not all genes in such networks are labeled with the function they perform. In this project, a two-phase approach was developed for the problem of function prediction - feature extraction from a partially labeled network followed by classification. The approach is robust to the network structure and was shown to dominate existing approaches.

For details, take a look at my paper.

Graph indexing (past)

This project dealt with indexing of a collection of relatively small graphs. For details, see here

Motifs in gene networks (past)

This project aimed at mining for motifs in a single large graph. For details, see here