Informatics Newsletter May 2018

Issue 15 of our School newsletter for students and staff.

New Staff

Academic

Petros Wallden - Lecturer in Security and Privacy - started on 1^st May 2018

Professional Services

Jodie Cameron - Support Office Secretary - started on 1^st May 2018

Research

Leimin Tain - Research Associate - started on 1^st May 2018
Hieu Hoang - Research Associate - started on 1^st May 2018
Martin Asenov - Research Assistant - started on 1^st May 2018
Chie Takahashi - Research Associate - started on 9^th May 2018
Chad Nester - Research Assistant - started on 17^th May 2018
Wouter Wolfslag - Research Associate - started on 8^th May 2018
Julie-Anne Meaney - Research Assistant - started on 1^st May 2018

Informatics Historical Trivia

This was then (in the 80s), this is now

Photos of Informatics, then and now — Top: examples of robotic arms | Middle: examples of chips | Bottom: students working.

Research successes

Wenfei Fan, professor of Web Data Management from the School of Informatics was elected as a Fellow of the Royal Society among fifty other eminent scientists on 9th May.

Wenfei is a computer scientist who has made fundamental contributions to both theory and practice of data management. He has both formalised the problems of querying big data and has developed radically new techniques that overcome the limits associated with conventional database systems.

His work has been adopted in the telecommunications industry for analysing massive data sets that defied their current technology. In addition, Wenfei has made seminal contributions to data quality, in which he devised new techniques for data cleaning that have found wide commercial adoption.

He has also contributed to our understanding of semi-structured data. The Royal Society is the oldest national scientific society in the world. It has a number of aims: promoting science and its benefits, recognising excellence in science, supporting outstanding science, providing scientific advice for policy, fostering international and global co-operation, education and public engagement. Each year the Royal Society elects up to 52 new fellows who join a cohort of the around 1,700 of the world's most influential scientists.

Michael Rovatsos was granted funding for hisr ReEnTrust: Rebuilding and Enhancing Trust in Algorithms project.

ReEnTrust will develop methods to help online Web-based platforms regain user trust by allowing users to evaluate and critique the algorithms used by these platforms, and to engage in constructive user-platform dialogue in order to recover from this loss of trust. As data-driven AI algorithms are starting to exert a massive influence on society and we witness breakdowns of trust increasingly often, new technological tools that allow users to explore the behaviour of algorithms and support the resolution of conflicts among users and platforms have the potential to greatly improve trust in data-driven AI. The ReEnTrust project, led by the University of Oxford in collaboration with the Universities of Edinburgh and Nottingham, aims to break new ground in this direction by producing novel algorithm analysis and mediation tools, and to develop a comprehensive responsible research and innovation methodology for the design of systems that embed these tools through extensive ethnography, stakeholder engagement, and participatory design activities.

The Lift project, lead by Dr Christophe Dubach at Edinburgh University, aims at simplifying the development of high performance computing algorithms for parallel processors commonly found in supercomputers and data centres.

Christophe's team has already received several awards for this work, including a best paper award at CGO 2018, the top conference in the field of compilers. Together with his colleague Dr Aaron Smith, Christophe received a MSR PhD scholarship funding and a Facebook research award to pursue research with Lift in the direction of accelerating neural networks using reconfigurable devices. Neural networks are currently the state of the art machine learning techniques used for imagine recognition or human languages translation. However, they require enormous computing resources. Reconfigurable devices offer the unique opportunity to specialise the hardware to the specifics of the application, leading to massive energy efficiency gain and large performance improvement. The goal is to derive, fully automatically, implementations of common machine learning algorithms at a fraction of the time required by human developers. The automation of this process will enable the exploration of a larger design space than currently feasible, leading to faster and more energy-efficient designs.

Lift project website

Jacques Fleuriot secured funding for the Digitizing industrial workflow project. Digiflow is an Industry 4.0 project funded by EIT Digital under its Digital Industry Action Line.

It revolves around a combination of IoT sensors, cloud infrastructure and workflow technologies to enable manufacturing companies to better monitor their shop floors and provide decision support towards efficiency optimisation. Partners include CREATE-NET from the Fondazione Bruno Kessler (FBK) in Italy, technology innovation company Reply Santer, Italian SME ThinkINside, and the School of Informatics’ WorkflowFM team consisting of Jacques Fleuriot and Petros Papapanagiotou. This project allows a unique combination of state-of-the-art technologies and skills brought together by each partner: (1) fog and cloud infrastructures for the management of IoT devices and the orchestration of associated processes by FBK, (2) expertise in product and business development for technological solutions with a worldwide reach by Reply Santer, (3) novel sensor technologies to track and analyze the activity of assets in a manufacturing floor by ThinkINside, and (4) the latest in workflow modelling and management research, including performance analytics, by the University of Edinburgh. The end-goal is a product that provides real-time monitoring, helps determine root causes of delays and inefficiencies, improves throughput, and reduces operational costs. The technology will be commercialised by Reply Santer. This is a key opportunity to make use of some of the latest research from the School of Informatics in a real, deployed solution that has the potential to create significant impact within manufacturing.

Researchers from Institute for Computing Systems Architecture had a number of papers accepted for major conferences:

- Two papers from ICSA researchers have been accepted for presentation at The 45th International Symposium on Computer Architecture (ISCA), the premier venue for computer architecture research: ProtoGen: Automatically Generating Directory Cache Coherence Protocols from Atomic Specifications automates the design of cache coherence protocols.

- In addition, Pramod Bhatotia and Myungjin Lee’s paper "ApproxIoT: Approximate Analytics for Edge Computing" has been accepted to appear at 38th IEEE International Conference on Distributed Computing Systems (ICDCS'18). ICDCS provides a forum for engineers and scientists in academia, industry, and government to present their latest research findings in major and emerging areas of distributed computing. The research was conducted with collaborators at Bell Labs. Pramod Bhatotia and the Cntr group’s paper "Cntr: Lightweight OS Containers" has been accepted for publications at USENIX ATC'18. The USENIX Annual Technical Conference brings together leading systems researchers for cutting-edge systems research and unlimited opportunities to gain insight into a variety of must-know topics, including virtualization, system and network management and troubleshooting, cloud and edge computing, security, privacy, and trust, mobile and wireless.

Researcher from the Centre for Intelligent Systems and their Application (CISA) Francisco J. Quesada Real and his team have been awarded the second prize at the Open IoT Challenge organised by the Eclipse Foundation.

The team (Francisco J. Estrella (PhD), Francisco Moya (Wave App) and Francisco J. Quesada Real (CISA, University of Edinburgh)) developed an idea of Trusting IoT. To do so, the group presented a Proof of Concept of a trusted positioning system for nursing homes that can be used to guarantee appropriate care of the elderly. The solution uses Eclipse Kura and Bluetooth low-energy beacons to implement the indoor positioning system. The most original and significant feature of their proposal is the use of the IOTA distributed ledger as a way to store a permanent and unalterable history of the positions of the residents at any point in time.

Deyi Xiong (Professor, Soochow University) and Rico Sennrich (Lecturer, School of Informatics) have been awarded a Royal Society Advanced Newton Fellowship to collaborate on Discourse-Level Neural Machine Translation. The quality of machine translation has improved markedly over the last few years, but the dominant approach is still to translate sentences in isolation. This ignores critical links (discourse is the study of such links) to the sentences that came before it. Translating sentences without context makes any form of coherence accidental and potentially misleading.

Advancing sentence-level NMT towards discourse-level NMT requires understanding the differences in discourse between different languages and pinpoint the most important discourse factors for translation. The project will exploit discourse annotation of parallel corpora in English and Chinese. Bonnie Webber (Professor Emeritus, School of Informatics) will be working with Deyi Xiong and his team at Soochow University on producing the corpus annotation. The primary result of the project will be a discourse-level neural machine translation system that advances translation quality by taking discourse-context into account in order to generate cohesive and coherent target texts that are easy to read and understand. This is also to mimic the way that human beings translate a text with the entire structure of it in their minds.

Funding

Recent grant applications table now available.

Grant Awards - May 2018 (77.68 KB PDF)

News from around the School

Scientists aid developments of warehouse robots

News stories - 'Scientists aid developments of warehouse robots'

Student news

Edinburgh University Formula Student (EUFS)

Edinburgh University Formula Student (EUFS) is a team of students dedicated to designing, manufacturing and building exciting race cars to compete at the world’s largest engineering competition - Formula Student. Running for their fourth year, EUFS has seen great success at competitions and have engaged with more than 300 students across the University. This year EUFS founded the new Autonomous branch, delving into the field of Informatics. The project seeks to transform one of the existing race cars into a completely driverless one, which must be able to drive around any race course without prior data. To achieve this the team will utilise/incorporate the fields of robotics, machine learning, control theory and parallel computing to develop an AI capable of surpassing human performance. There are no limitations on who can join the team as long as they are motivated and ambitious. Apart from technically minded students, EUFS is also looking for people interested in business and marketing.

If you are interested in helping you can get in touch with them via the link below.

Contact EUFS

The University of Edinburgh Hyperloop Team (HYPED)

The University of Edinburgh Hyperloop Team (aka HYPED) will be competing in the third round of the SpaceX Hyperloop Pod Competition this July. Building on last year's success, the team have engineered their second Hyperloop pod in order to demonstrate some of the key technologies of the future transportation, which was introduced and popularized by Elon Musk. HYPED is a society of around 80 students from all disciplines, including those from the School of Informatics, and we're looking forward to launching 'Poddy the Second' in California this summer.

HYPED website

EdIntelligence

EdIntelligence were awarded funding from IIG (Innovation Initiative Grants) to organise Artificial Intelligence Career Fair, The fair aims at providing more opportunities for School of Informatics masters and PhD students. It will feature jobs and internships for both groups in Natural Language Processing, Machine and Deep Learning and Vision. With about 30 companies, including Microsoft and Facebook attending the organisers hope that everyone in the field can benefit from the event. The companies involved will also be giving technical talks throughout the day and students will have the opportunity to have on the spot interviews as well as send their CVs to selected companies in advance.

Free registration for students will open next week on the Edintelligence's Facebook group.

EdIntelligence Facebook group

Spring 2018 grant awards page

Outreach and Public Engagement

Call for contributors, Midlothian Science Festival Schools Programme, deadline: 2nd July

The Midlothian Science Festival Schools Programme runs from Monday 8th to Friday 12th October. If you'd like to be involved through providing an inspirational activity for either primary- or secondary-school groups, please fill in the online form (link below) with details of your workshop/session. The deadline for proposals is 5pm on Monday 2nd July.

If you would like to contribute to the 2018 Midlothian Science Festival Schools Programme, please fill in the form via the link below.

Fill in the form

If you have any queries, please direct these to Jayne Quoiani.

Contact Jayne

ScotPEN: Scottish Public Engagement Network

Started in 2017, ScotPEN is a network of individuals (working within universities, the cultural sector, public bodies, etc) who support public engagement with research in various ways. The aims of the network are to share good practice and challenges and act as a central point of contact for funders.

More information about ScotPEN

To become involved in the network, if you are not already, then please visit this page via the link below.

Get involved

Save the date! - Explorathon 2018

Some of Scotland’s universities, including all of those in Edinburgh, will be taking part in European Researchers’ Night – Explorathon in Scotland – for the fifth year running. Do you fancy sharing your research at an event designed to reach over 3,000 people? Apply to our call for entries – coming soon!

Save the Date - Explorathon 2018

Events

A few spaces are available to attend the masterclass by Prof David B. Dunson, Duke University and Carnegie Cetenerary Professor. He is an expert in Bayesian methods, biomedical applications, machine learning and data science. His work is strongly motivated by applications ranging from genomics, ecology, neurosciences and other scientific fields to signal processing problems arising in restoring paintings and detecting forgeries. The Masterclass will comprise Lectures by Professor Dunson on Scalable Bayesian Inference. The Masterclass runs on 6th June 2018 2-5.30pm and 7th June 2018 9.30am-1pm

Details about the Masterclass

Contact Sally Galloway by Monday 4th June if you would like to attend.

Contact Sally

Michael Gutmann co-organises a workshop ABC in Edinburgh, as part of 2018 ISBA World Meeting. Approximate Bayesian Computation (ABC) performs inference when evaluating the likelihood function is infeasible but sampling from the model is possible. First developed in population genetics, ABC is now a well-established technique in this and other diverse disciplines including infectious disease epidemiology, astrophysics and econometrics. The ABC in Edinburgh workshop will feature invited and contributed talks, and will have a poster session to further exchange and discuss recent work on the theoretical underpinnings, computational advances, and applications of approximate Bayesian computation.

ABC in Edinburgh

Short courses