Gayathri Nadarajan

Centre for Intelligent Systems and their Applications
School of Informatics
University of Edinburgh
2.02 Informatics Forum
10 Crichton Street
Edinburgh EH8 9AB, Scotland
United Kingdom
E-mail: gaya.n [at]

Research  Teaching  Publications  Other

I am now an assistant professor at Sungkyunkwan University (SKKU). I can still be reached at the email address above or at gaya [at]

I teach data science courses and am enthusiastic in applying intelligent techniques on big data such as videos and more recently, medical data, in particular, drug-related data.

Current Research Interests: Biomedical Data Integration and Analysis, Predictive Maintenance, Computational Drug Repurposing, Graph Databases, Biomedical Ontologies.

Previous Research Interests: Automatic Workflow Composition, HTN Planning, Ontology Engineering, Job Scheduling and Monitoring, Process Modelling, Image Processing Tools.


At Informatics, Edinburgh

I was a researcher with the Fish4Knowledge (F4K) project (2010-2013). My roles included:

I obtained a Ph.D. in 2010 with my thesis Semantics and Planning Based Workflow Composition for Video Processing. I worked with Jessica Chen-Burger and Bob Fisher and continued to do so in F4K. My research involved investigating methods to model generic vision tasks and to provide an automatic mechanism to allow users without image processing expertise to perform complex vision tasks. Here is a brief statement of my work:

Much of video analyses are conducted manually. Automatic solutions are only possible with image processing expertise and access to specialised tools. Furthermore, each time a task is required to be done, image processing experts develop the applications from scratch and do not reuse existing modules that they have implemented previously due to the inherent difficulties in estimating the input, output and threshold values. Workflow systems, on the other hand, have several drawbacks including construction of workflows automatically. I propose a hybrid workflow composition approach that will enable the selection of image processing tools based on their performance levels automatically or semi-automatically (with user involvement). In order to achieve this ontologies are used to assist with semantic interoperability and planning to assist with performance-based selection. On a wider context, the solution I propose will enable image processing-naive users, such as domain experts to conduct complex video processing tasks. It would also provide a reasonable automatic solution for various video analyses that would take too long to perform manually. The use of semantics-based technologies would also enable this system to be integrated within a distributed environment, such as the Grid. Read full proposal here. Slides.

I visited the National Center for High Performance Computing (NCHC), NARL, Taiwan in summer 2012 and spring 2013 to collaborate with HPC experts on accessing and managing their facilities for F4K.

I visited University of Catania, Italy between September and December 2008 to collaborate with image processing experts on modularisation methodologies.

I was involved with the OpenKnowledge project in 2008 and the SynProc project in 2009-10.

My current and past collaborators include Dave Robertson, Omer Rana (Cardiff), Concetto Spampinato (Catania, Italy), Rafael Tolosana (Zaragoza, Spain), Fang-Pang Lin (NCHC, Taiwan), Arnaud Renouf (GREYC, France) and James Malone (EBI, Cambridge).


2017 - 2018 AI consultant, team lead and instructor for various institutions.

2016 - 2017 Senior Scientist, Standigm, South Korea.

2014 - 2015 Senior Researcher, Seoul National University, South Korea.

2010 - 2013 Research Associate, University of Edinburgh, UK.

2005 - 2010 Ph.D. in Informatics, University of Edinburgh, UK.
Thesis: Semantics and Planning Based Worklflow Composition and Execution for Video Processing.

2004 - 2005 MSc. in Artificial Intelligence, University of Edinburgh, UK.
MSc. Project: Translating FBPML to a Semantic Web-Based Language.

2002 - 2004 Lecturing and tutoring at University Tenaga Nasional, Malaysia.
Courses: Software Engineering, C Programming, Artificial Intelligence, Computer Literacy, Operating Systems.

1999 - 2002 Software Engineer with Altion Ltd in Dublin, Ireland.
Software development mainly involving Perl, C and Java for network telecommunications software.

1995 - 1999 Completed undergraduate in Computer Science in Trinity College Dublin, Ireland.
Final year project: On-line Handwriting Recognition using Hidden Markov Models.


2009 & 2010: System Design Project
2008: InfBase: Support for Learning
2004 - 2009: Informatics 1B (Object-Oriented Programming, Data and Analysis)
2004 - 2008: Informatics 1A (Functional Programming, Computation and Logic)
2006 - 2008: Software Engineering with Objects and Components
2006 - 2007: Informatics Research Review



Heriot-Watt University, Jan 2016.
Seoul National University, Sept 2014.
SEAP'13 Workshop, Taiwan. Invited Talk, Dec 2014.
NCHC, Taiwan. Invited talks, July 2012 (Taichung), April 2013 (Hsinchu).
Konkuk University, Seoul, Korea. Invited talks, Sept & Dec 2012.
3DPAS Workshop. eScience Centre, Edinburgh, April 2011.
SICSA DEMOfest poster, November 2009.
Poster at Google Retreat, Zurich, July 2009.
Informatics Jamboree Poster, June 2007.
Software Systems and Processes Group (SSP) Talk, Oct 2006.
IDIAP 15th Anniversary Workshop Invited Talk. Martigny, Switzerland, Sept 12th-13th, 2006.

Services & Personal

Tae Kwon Do, Korean martial art.
Social Events Convener with the Informatics Graduate School (2006-10). Highlights: Firbush Trip (with photos), Death by Dessert championship, Go-Ape Forest adventure, Paintball (Photo by Petros), Hillwalking in Pitlochry (Photos by Maria Luisa). Please contact Alex/Iain if you have any interesting event in mind!
CompuCast, a podcast for computer scientists by Informatics, Edinburgh University.
Some Personal Photos.