Seminar

Evaluation of Network Resilience and Survivability: Analysis, Simulation, Tools, and Experimentation by James P. G. Sterbenz, ITTC, University of Kansas

As the Internet becomes increasingly important to all aspects of society, the consequences of disruption are increasingly severe. Thus it is critical to increase the resilience and survivability of the future networks in general, and the Internet in particular. We define resilience as the ability of the network to provide desired service even when the network is challenged by attacks, large-scale disasters, and other failures. Resilience subsumes the disciplines of survivability, fault-tolerance, disruption-tolerance, traffic-tolerance, dependability, performability, and security. After an introduction to the disciplines and challenges to network resilience, this presentation will discuss analytical, simulation, and experimental emulation techniques for understanding, evaluating, and improving the resilience of the Future Internet. This includes a multilevel state-space based approach that plots network service delivery against operational state that is the basis for both mathematical- and simulation-based analysis, and graph-theoretic complex-system approaches that embed fundamental properties such as redundancy and diversity into all aspects of network structure, mechanism, and protocols. A set of tools to help in this analysis has been developed: KU-LoCGen (Location and Cost-Constrained Topology Generation), KU-TopView (Topology Viewer), and KU-CSM (Challenge Simulation Module). Plans to experimentally evaluate resilience include using the international programmable testbed GpENI: Great Plains Environment for Network Innovation. A new composable, cross-layered resilient transport protocol (ResTP) and geodiverse multipath routing protocol (GeoDivRP) are being developed.

Event details

  • When: 28th January 2014 14:00 - 15:00
  • Where: Maths Theatre B
  • Series: School Seminar Series
  • Format: Seminar

Techniques for Scalable Privacy-Preserving Record Linkage

Three-Hour Research Tutorial by Peter Christen, Australian National University

Privacy-Preserving Record Linkage (PPRL) is an increasingly important topic in data management, data engineering, and data mining, as organizations in both the private and public sectors are under pressure to share, integrate, and link their data in order to allow analysis that is not possible on individual databases. At the same time, sensitive information such as personal identifying details or confidential business data need to be protected. PPRL can for example be applied to match health databases without revealing any sensitive personal details of patients, or to detect individuals that have been involved in fraudulent activities across organizations without the need to share the full, potentially confidential, databases. Research in PPRL over the past decade has developed a variety of algorithms, however the challenge of linking very large databases in privacy-preserving, scalable, accurate, and automatic ways is still an open problem. In this half-day tutorial I will illustrate the significance of PPRL through several real-world scenarios, and introduce the concepts, techniques, algorithms, and research directions of PPRL.

This tutorial was previously held at CIKM 2013. Further details at: http://cs.anu.edu.au/people/Peter.Christen/cikm2013pprl-tutorial/.

There will be a limit on numbers due to the size of the venue. Attendance is free but prior registration is required. Please contact Graham Kirby <graham.kirby@st-andrews.ac.uk> if you wish to attend.

Event details

  • When: 11th February 2014 14:00 - 17:00
  • Where: Cole 1.33a
  • Format: Seminar

Intellectual Property – What, why, and how?

Dr Alan Boyd and Dr Nicholas Malden, D Young & Co LLP (Patent Attorneys)
Intellectual Property – What, why, and how?
In many technological fields, especially computer science, output isn’t a physical object that can be sold, but is an idea or an expression of an idea that comes from your mind. But how can industry take advantage of such ideas or expressions, and what’s to stop a competitor from simply copying your hard work and benefiting from it at your expense? In this talk, we’ll take you through the basics of Intellectual Property (IP) and look at how it can be used by a computer scientist, its importance to new start-ups and large multinationals alike, and also introduce you to the patent attorney profession. We’ll give you a chance to try your hand at writing a patent claim, and we’ll give a prize to the best one!

Event details

  • When: 5th November 2013 14:00 - 15:00
  • Where: Purdie Theatre C
  • Series: School Seminar Series
  • Format: Seminar

Some Limits of Language: A Perspective from Formal Grammars and Languages by Prof Arvind Joshi, School of Engineering and Applied Science, University of Pennsylvania, USA

Limits of language can be studied from various perspectives such as morphology, syntax, and semantics, among others. At the syntactic level, one direction that has been pursued very actively is via the theory of formal languages, beginning with the so-called Chomsky hierarchy. In this hierarchy, the finite state languages (regular languages) and the context free languages (CFL) have been studied very extensively, leading to many important results relevant to syntax as well as parsing. At the same time, inadequacy of these two classes of grammars (languages) for capturing natural languages has been well documented by now.
A careful look at the pumping lemma for context free languages led to the discovery of the so-called tree adjoining grammars (TAG) and to the notion of mildly context sensitive languages (MCSL), which has served as a framework for showing equivalences among other systems (such as Combinatory Categorial Grammars, CCG and Head Grammars (HG), for example). This has led to a deeper understanding of the limits of language, more specifically, by trying to provide an answer to the question: how far do we need to be beyond CFL to achieve syntactic adequacy.
I will try to describe some of this recent work by a number of researchers in the past few years.

Aravind Joshi did his undergraduate work in India and his graduate work at the University of Pennsylvania in Electrical Engineering, while simultaneously attending some courses in Linguistics at the same University. Since 1961 he has been a faculty member in the Department of Computer and Information Science and the Department of Linguistics. At present, he is a Professor of Computer and Cognitive Science and a Member of the Institute for Research in Cognitive Science.
Besides working on some problems in the mathematics of language, at present, he is involved in a project on discourse annotation, jointly with Bonnie Webber (Edinburgh U.) and Rashmi Prasad (U. Wisconsin), for creating the Penn Discourse Treebank (PDTB).

Event details

  • When: 15th October 2013 14:00 - 15:00
  • Where: Purdie Theatre C
  • Series: School Seminar Series
  • Format: Seminar

St Andrews Students – Experience with Industrial Internships

Internships are fantastic opportunities to gain some practical experience as well as find out what is happening the real world of computer science! Come and hear some our UG students share their experiences of their 2013 summer internships.
Melissa Mozifian: Adobe
Waqas Arshad: AIG
Mariya Hristova: Google STEP
Sam Koch: Facebook

Event details

  • When: 1st October 2013 14:00 - 15:00
  • Where: Purdie Theatre C
  • Series: School Seminar Series
  • Format: Seminar

Scaling Skyscanner’s Flight Search & Making Mobile Applications at Skyscanner

Welcome to the first presentation in the School of Computer Science’s Seminar Series.

Please join us on for a seminar on Skyscanner’s technology this Tuesday (September 24) at 14:00 in Purdie Lecture Theatre C.

The two talks combined will take about 50 minutes with time for questions, and combine two topics presented by Skyscanner experts:

Grzegorz Janas- Project Manager Mobile Applications
Simon Thorogood- Senior Architect Development Engineering
Scott Krueger- Technical Manager Databases
Skyscanner: http://www.skyscanner.net/

Abstract:
Come and hear from Skyscanner Engineers on the Challenges behind engineering the world’s fastest growing Metasearch product and our journey towards being the “most trusted online travel company in the world” Skyscanner has 24 million unique monthly visitors and 25 million installed apps generating one-third of its traffic through mobile.

Everybody welcome.

Event details

  • When: 24th September 2013 14:00 - 15:30
  • Where: Purdie Theatre C
  • Series: School Seminar Series
  • Format: Seminar

On Normalising Disjunctive Intermediate Logics

Speaker:
Prof. Jonathan Seldin, University of Lethbridge, Canada

Abstract:
In this talk it is shown that every intermediate logic obtained from intuitionistic logic by adding a disjunction can be normalized. However, the normalisation procedure is not as complete as that for intuitionistic and minimal logic because some results which usually follow from normalisation fail, including the separation property and the subformula property.

Biography:
Jonathan P. Seldin, now Professor Emeritus, is a well-established senior scientist at the University of Lethbridge, Alberta, Canada, with an Amsterdam PhD in combinatory logic supervised by Haskell Curry. This logic, together with lambda-calculus (to which it is equivalent) is a prototype for functional languages, such as Haskell, and typed lambda-calculus is a prototype for the typing discipline in programming languages. His work on lambda-calculus, both pure and typed, has applications in formal verification, the use of formal logics to prove properties of programs (e.g., that they satisfy their specifications). He has co-authored works with Curry and Hindley on combinatory logic and lambda calculus. He is also interested in the history and philosophy of mathematics and in proof normalisation and cut-elimination for various systems of formal logic. His visit to Scotland is as a SICSA Distinguished Visiting Fellow, to work with Prof. Kamareddine at Heriot-Watt University and with Dr Dyckhoff at St Andrews. For details and publications see http://directory.uleth.ca/users/jonathan.seldin

Event details

  • When: 3rd September 2013 11:30 - 12:30
  • Where: Cole 1.33a
  • Format: Seminar

SACHI Seminar: Team-buddy: investigating a long-lived robot companion

SACHI seminar

Title: Team-buddy: investigating a long-lived robot companion

Speaker: Ruth Aylett, Heriot-Watt University, Edinburgh

Abstract:
In the EU-funded LIREC project, finishing last year, Heriot-Watt University investigated how a long-lived multi-embodied (robot, graphical) companion might be incorporated into a work-environment as a team buddy, running a final continuous three-week study. This talk gives an overview of the technology issues and some of the surprises from various user-studies.

Bio:
Ruth Aylett is Professor of Computer Sciences in the School of Mathematical and Computer Science at Heriot-Watt University. She researches intelligent graphical characters, affective agent models, human-robot interaction, and interactive narrative. She was a founder of the International Conference on Intelligent Virtual Agents and was a partner in the large HRI project LIREC – see lirec.eu. She has more than 200 publications – book chapters, journals, and refereed conferences and coordinates the Autonomous affective Agents group at Heriot-Watt University- see here

Event details

  • When: 10th September 2013 13:00 - 14:00
  • Where: Cole 1.33a
  • Format: Seminar

Jacob Eisenstein: Interactive Topic Visualization for Exploratory Text Analysis

Abstract:
Large text document collections are increasingly important in a variety of domains; examples of such collections include news articles, streaming social media, scientific research papers, and digitized literary documents. Existing methods for searching and exploring these collections focus on surface-level matches to user queries, ignoring higher-level thematic structure. Probabilistic topic models are a machine learning technique for finding themes that recur across a corpus, but there has been little work on how they can support end users in exploratory analysis. In this talk I will survey the topic modeling literature and describe our ongoing work on using topic models to support digital humanities research. In the second half of the talk, I will describe TopicViz, an interactive environment that combines traditional search and citation-graph exploration with a dust-and-magnet layout that links documents to the latent themes discovered by the topic model.
This work is in collaboration with:
Polo Chau, Jaegul Choo, Niki Kittur, Chang-Hyun Lee, Lauren Klein, Jarek Rossignac, Haesun Park, Eric P. Xing, and Tina Zhou

Bio:
Jacob Eisenstein is an Assistant Professor in the School of Interactive Computing at Georgia Tech. He works on statistical natural language processing, focusing on social media analysis, discourse, and latent variable models. Jacob was a Postdoctoral researcher at Carnegie Mellon and the University of Illinois. He completed his Ph.D. at MIT in 2008, winning the George M. Sprowls dissertation award.

Event details

  • When: 23rd July 2013 13:00 - 14:00
  • Where: Cole 1.33
  • Format: Seminar

System Seminar: Middleware support for wireless sensor network, by Prof. Danny Hughes, KU Leuven, Belgium

Abstract:

Contemporary ICT infrastructures are trending towards a pervasive substrate of internet-connected sensors, actuators and human interfaces. Effective use of this pervasive infrastructure is key to solving 21st century challenges such as: mass transportation, energy conservation and environmental monitoring. Building effective applications that execute on this infrastructure requires advanced middleware support that respects the resource constraints of embedded devices. In this talk, I will present recent work from KU Leuven in the area of middleware support for sensing applications, with a focus on the Loosely-coupled Component Infrastructure (LooCI). I will then discuss emerging opportunities for ‘social sensing’ wherein online social networks are used to source and configure participatory sensor networks.

Bio:

Dr. Danny Hughes is an Assistant Professor with the iMinds-DistriNet group of the department of Computer Science at KU Leuven, Belgium. He received his PhD in Computer Science from Lancaster University in 2007. He has since worked as a Visiting Scholar at UC Berkeley (USA) and Universidade de São Paulo (Brazil) and as a lecturer at Xi’an Jiaotong-Liverpool University (China). His research interests are in distributed systems, with a focus on middleware support for dynamic systems such as wireless sensor networks, peer-to-peer networks and online social networks. You can find out more about his research at: https://distrinet.cs.kuleuven.be/people/showMember.do?memberID=u0061846

Event details

  • When: 2nd May 2013 14:00 - 15:00
  • Where: Cole 1.33a
  • Format: Seminar