Parallelisation of Network Algorithms
Complex networks have received a lot of attention in recent years. One interesting source of such networks are social networking applications such as Twitter and Facebook, from which friendship networks can be extracted, where nodes represent users of the social networking service and links between nodes represent social connections. Businesses are interested in exploiting such networks in order to target personalised services. To support this, a number of algorithms have been developed to extract communities from social networks and analyse information flow through networks. These algorithms need to be able to work on very large-scale networks, as social networks can contain millions of nodes with tens of millions of links between them. This project will involve the use of high-performance computing to scale network analysis algorithms to big datasets. A number of different high performance platforms are available. The precise analysis and platform to be used will be studied in consultation with the student.