Pol Mac Aonghusa

School of Computer Science & Statistics

Pol Mac Aonghusa

macaongp at scss dot tcd dot ie
School of Computer Science and Statistics
Room 140, Lloyd Institute,
Trinity College
Dublin 2, Ireland

You can also reach me at my IBM Research email:

aonghusa at ie dot ibm dot com

Research

Data Privacy

My main research interest is in Data Privacy - primarily from a consumer perspective.

As the Infographic to the right shows, there is no universal - one size fits all - view of privacy shared by all users.

I believe that online privacy is a pragmatic tradeoff between perceptions of utility, risk and cost. Users are preapred to trade an acceptable amount of risk in return for utility gained. So rather than adopting a traditional minimal-disclosure approach, how can we enable the user to make informed, balanced choices. I am interested in real and practical techniques the user to manage their privacy risk - rather than a traditional hiding or anonymisation approach. I'm currently looking at how users can actively detect, understand and adapt their privacy utility-risk-cost balance as they interact with systems. While it may seem like the commerical big guys hold all the cards, the good news for the user is there are plenty of free, large-scale, online laboratories where we can test ideas - online search and recommender systems.

This is a relatively new research area - investigating methods for privacy as a (user) serviceml xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns:mv="http://macVmlSchemaUri" xmlns="http://www.w3.org/TR/REC-html40">

Pol Mac Aonghusa

macaongp at scss dot tcd dot ie
School of Computer Science and Statistics
Room 140, Lloyd Institute,
Trinity College
Dublin 2, Ireland

You can also reach me at my IBM Research email:

aonghusa at ie dot ibm dot com

Research

Data Privacy

My main research interest is in Data Privacy - primarily from a consumer perspective.

As the Infographic to the right shows, there is no universal - one size fits all - view of privacy shared by all users.

I believe that online privacy is a pragmatic tradeoff between perceptions of utility, risk and cost. Users are preapred to trade an acceptable amount of risk in return for utility gained. So rather than adopting a traditional minimal-disclosure approach, how can we enable the user to make informed, balanced choices. I am interested in real and practical techniques the user to manage their privacy risk - rather than a traditional hiding or anonymisation approach. I'm currently looking at how users can actively detect, understand and adapt their privacy utility-risk-cost balance as they interact with systems. While it may seem like the commerical big guys hold all the cards, the good news for the user is there are plenty of free, large-scale, online laboratories where we can test ideas - online search and recommender systems.

This is a relatively new research area - investigating methods for privacy as a (user) service. We merge techniques from Data Mining, Machine Learning, together with mathematical frameworks for privacy such as Bayesian, Information Theoretic and Differential Privacy Models to give the user more control of their privacy.

Applied Data Privacy is also a practical research topic for my work at IBM. Complex and increasingly connected systems only increase the challenges for data publishers. It's clear there is a need for new thinking and technologies. I co-authored a high-level article a while back in the IBM Journal of Research and Development called Privacy protection in open information management platforms starting to think about some of the challenges. It's something I'd like to explore further.

Infographic source: www.thedrum.com.

 

Computer Science
I work at IBM Research's Dublin Laboratory where I lead research teams applying semantics for large-scale cataloguing, processing and integration of semi-structured data.
Team research interests span data management for semi-structured data, parallel methods for data intensive processing, Semantic Web, Linked Data, reasoning with Web data, flexible data integration methods, stream processing, peer-to-peer and other distributed systems.
Our use cases are mostly drawn from urban systems and more recently from person-centric care (Social care and Healthcare in particular). The team has won second prize at the Semantic Web Challenge in 2012 and third prize in 2013 for this work.
It's probably easiest to explain if you check out some of our projects on the IBM Research site:

Papers

External Projects

A big part of my IBM Research job team is participation in EU Research programs. We have proposals underway in H2020 FET, ICT and Societal Challenge tracks.

We are partners in two FP7 projects at the moment:

Links to Interesting Resources

I suggest installing the Mozilla add-on LightBeam and using it to explore third party web tracking. It’s not something you want to keep running constantly as it does sometimes slow down the browser for sites with many third party cookies. But as a simple intereactive tool to inform yourself it is an excellent start.

 

Me & My Shadow (myshadow.org) is an interesting list of user privacy tools. A lot of design has gone into the site to make it attractive and simple to navigate.

 

While you are on Me and My Shadow you should check out Trackography. It’s a great way to understand all that third party tracking that goes on. Just follow the instructions. Again hats off to the team working on this as they have clearly invested a lot of work in making a really usable and interesting tool.