carry out research in the field of Algorithms with a keen interest in Distributed Computing. Currently research interests revolve around Network Searchig, Network Patroling, Agent Synchronisation and Randomised Algorithms. Recenly, I have also had a growing interest in the area of Social Network Analysis. Use the links on the left to browse through the projects I have been involved with so far. Keep checking this space for information about new projects I am working on.

study a randomised distributed communication-less coordination mechanism for uniform anonymous agents located on an undirected ring. When woken up the agents have no coherent sense of direction. The agents perform their actions in synchronised rounds. The main task to be performed by each agent is to determine the initial position of every other agent and eventually to stop, or proceed to another task, at its initial position in a fully synchronised manner. Our main result constitutes of a fully distributed randomised algorithm that solves the location discovery problem with very high probability in O(n log2 n) rounds. Simulation software used to help understand this problem further is available to download below and a research paper relating to this topic can be found in the publications section of this site. My MSc dissertation is also available in the publications section and is closley related to this research.

Agent Patroling Simulation Software Download

the research that we are carrying out is based around social network analysis. We are looking at different aspects of these networks including level of degree for vertices as well as possible weighting methods we can employ to these vertices and the edges in the network to help find the most influential nodes in a network. We use datasets gathered from popular social networking sites. We have also been looking at Bottleneck identification and grouping based on interests amongst nodes. The types of social networks we are most interested in are Facebook and Twitter connections. Below are links to some of the images we created as part of our initial research into this area.

University of Liverpool Computer Science PhD Social Network

My Facebook Social Network

Social Network for Crosby, Liverpool (My Home Town)

Combined Social Network for Crosby and the Computer Science Department from the University of Liverpool (Update expected soon)

Full Social Network Graph for all Data Collected so far (Update expected soon)

Click Here to participate in this project. Simply log in using your Facebook account, wait for your friends list to be generated and then click "go" and wait for confirmation of your submission to our project.

is a project that we worked on over the course of the DEVxS 2011 conference. At the conference a competition was held for everone present that involved doing something interesting with libary data that was provided. Our entry took the data, cleaned it up, made it easily searchable and also provided a front end interface for the user. When searching the data, on books and publications, information is shown about what other items were read by the readers of the searched item. OokNog was also our team name. It is a Dutch word that means "Also" and so is an adept choice. The team consisted of Andrew Colins, Jude-Thaddeus Ojiaku, Arnoud Pastink and myself. We won first prize.

OokNog Website

is the most recent work I have been involved with. We consider a group search problem in which k MEs (mobile entities) located on the line perform search for a specific destination. The MEs are initially placed at the same point (origin) on the line L and the target is located at unknown distance d either to the left or to the right from the origin. The problem with k = 1 is known as the cow-path problem, and the complexity of this problem is known to be 9d in the worst case, it is also known that this is the case for k >= 1 MEs. In this paper we present a clear argument for this claim by showing a rather counter-intuitive result. This complexity also applies to any metric. Namely, in any metric independently from the number of MEs group search cannot be performed faster than in time 9d; where d is the distance between the origin and the destination. We also look at the case when there are k = 2 MEs with different speeds. It is assumed that the search is accomplished when all MEs eventually arrive at the destination.

[2] T. Friedetzky, L. Gasieniec, T. Gorry, and R. Martin, Observe and Remain Silent, Sent to 37th International Symposium on Mathematical Foundations of Computer Science, 2012.

[3] S. Das, L. Gasieniec, T. Gorry, R. Martin, M. Shalom, and S. Zaks, Group Search on the Line - A Tale of 9D, Submitted to Search and Rendezvous Workshop - Lorentz Center, 2012.

[1] T.Gorry, Distributed Agent Location Discovery on the Ring, MSc Dissertation, University of Liverpool, 2011.

I am currently demonstrating for COMP527: Data Mining. Here are some useful links for this module:

Dr Irina Biktasheva - Home Page

COMP527: Data Mining Lecture Notes

COMP527: Data Mining Assignments

COMP527: WEKA Website

COMP527: Google Search for .ARFF datasets for WEKA

Labs for this module are on Monday 10:00 to 12:00 and Friday 15:00 to 17:00 and demonstrators are myself and Richard Stocker.