Department of Computer Science

6th Annual Postgraduate Conference

in

Computer Science

20-21 March 2007
Department of Computer Science
Queen Mary University of London
Physics 602

The 6th Annual Postgraduate Conference will take the form of an informal 2-day workshop of talks with a poster session and discussion panels. This year the format of the conference is changed to give different options to the students according to their stage of research.

Official Schedule

Tue 20 March	Wed 21 March
10:00-10:30: tea & coffee	10:00-10:30: tea & coffee
10:30-10:45: introduction	10:30-11:45: session 4 Louise Valgerður Nickerson (IMC) Strategies for designing auditory overviews Hany Azzam (IR) Retrieval of large-scale semantic data: investigating a hybrid DB+IR approach Han Chen (Logic) Quantitative analysis of language in multi-threaded programs Jun Li (Vision) Image colorization with neighborhood matching Awais Ahmed Awan (NRL) Synergy: blending heterogeneous measurement elements for effective network monitoring
10:45-12:15: session 1 Peter Hearty (RADAR) Predicting something when you known nothing Caifeng (Kevin) Shan (Vision) Capturing canonical correlations for facial analysis George Papatzanis (IMC) Identifying phenotypes and genotypes: a case study on evaluating an in-car navigation system Jun Wang (IR) Informativeness of tf-idf and language modeling Milan A. Verma (Vision/IMC) On biologically-motivated saliency detection Dave Russell (Vision) Dynamic decomposition of scenes



	11:45-12:00: poster boaster Samuel Pachoud (Vision) Video sequence similarity
	12:00-12:45: lunch/poster session
12:15-13:15: lunch

	12:45-13:30: session 5 Corrado Biasi (Logic) A combinatorial proof of strong normalization for the simply typed lambda calculus with pairings and projections Bushra Akhtar (Vision) Quantification of displacement and deformation of chondrocytes under pressure Jean-Baptiste Thiebaut (IMC) The emergence of graphical music

13:15-14:30: session 2 Oussama Metatla (IMC) What makes an auditory diagram diagrammatic? Wei Li (NRL) Toward accurate efficient traffic classification Frederik Forst (IR) POLIS: a summarisation logic for structured documents Aziem Chawdhary (Logic) Proving Termination of Systems Code Chrystie Myketiak (IMC) Developing 'community' in an online chat environment
	13:30-13:45: break
	13:45-14:30: panel 2: Risk/Uncertainty in PhD project Projects inevitably involve risk and concerns about risk are regular worries of project managers. ’Risk Management’ has become an important part of ‘Project management’. It is even said that project management is nothing more than risk management. As PhD students we all face with different risks/uncertainty in our projects. Regardless of the topic, in our PhD project we are trying to achieve 1) a piece of research with a required level of quality (thesis and viva) 2) in a specific time (3-4 years). In order to achieve the project’s objectives we have to learn how to deal with risks and uncertainty in our project. In its easiest way, risk management has 3 main stage: Risk identification: What are risks and what are sources for them? How they affects our projects? Risk measurement: How can we analyse, measure and prioritise risks. Risk response: How can we response to mitigate risks or reduce their effects This panel aims to discuss about important risks that might happen in a PhD project and debate about more common risks and probably with some hints about how to deal with them.


14:30-14:45: break	14:30-14:45: closing & prizes
14:45-15:45: session 3 Christopher Frauenberger (IMC) The city of ears ))) - hearing is believing Arash Eshghi (IMC) Multi-party conversation: implications for dialogue models Taoyang Wu (Logic) The pseudo-random world
	15:00-16:00: DLS in CS446: Cryptography: The Art of Knowing Nothing


15:45-16:00: break
16:00-16:45: panel 1: Promoting Interaction within DCS PhD Community Most communication within the DCS PhD community seems to be superficial; you see people and say hello but little comes out of the communication that impacts your PhD. This should not be the case. The panel will discuss how building real relationships within the PhD community will directly impact your progress. Sharing experiences with fellow PhD students (even from different research groups) can be as helpful as meeting your supervisor. However, this type of support seems to only exist within rooms and at best across a research group. There is a wealth of knowledge and support that exists across the department that students can draw on. Chance meetings and social gatherings help build relationships across the department but it must be supplemented with new methods. The panel will propose an idea about setting up a trial on-line community (i..e, Facebook, Myspace), solely consisting of DCS PhD students. The idea is to create better integration and breakdown the barriers that seems to exist amongst a small group of people who are going through very similar experiences.

	18:30- : social gathering: - dinner at Rama Thai subsidised by the dept!!

This year we are introducing the following:

More feedback: All submissions will be peer reviewed by two people (on top of on the day feedback)
Poster session: For MSc and early stage PhD students (should they desire)
Panels: Given by late stage PhDs to share their experience and to be interspersed with the talk sessions
Official prizes!: For best submission (paper/poster) and presentation.
Publication: Of the best ten extended abstracts in a DCS technical report

Goal

The goal of these conferences is to strengthen links within our research community and to keep abreast on the interesting work going on in other groups within the department. Furthermore, it is an excellent opportunity to practice the skills involved in preparing and giving a talk in a relaxed and friendly atmosphere.

What's required

All registered MPhil and PhD students are expected to submit a two-page extended abstract of their work and present it in a 10 minute talk. Presentations will serve as an overview of PhD work up to date without going into technical detail. You will also need to peer review 2-3 other submissions.

Early stage or MSc?

Students in the very early stages of their research, will have the option to do a poster and will submit one-page abstract of the poster. Poster session will include "boaster session" so be prepared to say a few words.

Late stage?

Students in late stages have the option to organise a panel to share experience but they should submit a panel proposal including other panel members and a topic.

Peer review

All submissions will be peer reviewed by two others - one from the same group, one from another group. All presentations will also be reviewed by five other people. Forms will be provided for each set of reviews. This is to ensure that each person gets the maximum out of participating.

Prizes

This year, we will be presenting the following prizes:

Best abstract: to be determined by the peer review process.
Best poster: to be voted on by all.
Best presentation: to be voted on by all.
Best reviewer: based on feedback from reviewers.

Who else can attend?

All Queen Mary Computer science postgraduate students and research assistants are invited to come along if they can, and MSc students and research staff may also give talks if they wish. Students from other University of London sites and QM departments will be invited to attend although presentations will be limited to Queen Mary Computer Science students. Faculty members and supervisors will not be attending, to keep the atmosphere as relaxed as possible.

Coffee and a small buffet lunch will be provided for all presenters.

Deadlines

Please submit a title and extended abstract to: pg-conference@dcs.qmul.ac.uk by the following dates:

31 Jan: Abstract/Poster/Panel Title
15 Feb: Extended abstract

Your submission must include the following information:

Abstract	Poster	Panel
Name Year Research group Supervisor(s) Abstract title 2 page abstract	Name Year Research group Supervisor(s) Poster title 1 page abstract	Name Year Research group Supervisor(s) Panel topic Panel description Panel members