Battling big data

Academics from Aston University have received a grant totalling almost half a million pounds from the Technology Strategy Board to investigate big data exploration.

Researchers from the School of Engineering and Applied Science will work with leading industry specialists based at the Innovation Birmingham Campus; IGI Ltd and Daden Ltd, on a three-way collaborative project which could revolutionise how big data is analysed.

Big Data is something of a buzzword at present. It has many meanings, but typically the term relates to working with large, often heterogeneous, data sets, and in particular, analysing and interpreting these data sets to extract information.

Based at Faraday Wharf on the Innovation Birmingham Campus, IGI Ltd was established in 1983 as a geochemical consultancy. In past 30 years, IGI has grown to be the world’s largest independently owned geochemical consultancy company, and works with all the major, and many smaller, oil and gas exploration companies. Also based at Faraday Wharf, Daden Ltd was founded in 2004, and since 2006 has established itself as a leading UK and global company working in the area of immersive environments and artificial intelligence.

This three-way venture, between the Non-Linearity and Complexity Research Group in the School of Engineering and Applied Science at Aston, IGI Ltd and Daden Ltd, will address the problem of making big data exploration work, focusing on traceability and usability. Ian Nabney, Professor of Computer Science, will be leading the machine-learning aspects of the project, while Jo Lumsden, Senior Lecturer, will provide input on human factors.

The aim of this project is to develop systems that help non-specialists understand and analyse their data more easily through data visualisation: transforming high-dimensional data into a low-dimensional space that can be displayed in 2D or 3D graphical format.

IGI’s role will focus on data capture, cleansing and annotation; their aim is to develop a system to improve the efficiency and automation of the data import through cleansing and quality control processes. Daden will be responsible for scalable 3D immersive visualisation, providing research and development for a 3D rendering engine optimised for immersive 3D data display. Tying the partnership together, academics at Aston will be assisting in visualisation, focusing on scalability (i.e. making algorithms work efficiently with very large datasets), fidelity (providing feedback to the user on how well the visualisation reflects the actual structure of the data) and usability (ensuring that advanced algorithms can be used appropriately by non-specialist users).

The economic impact of this project is expected to be very significant, achieved through commercial exploitation of the innovations developed in the project by the two industrial partners.

Societal impacts will follow from the improved analysis and understanding of extremely large datasets and thus the improved decisions that will be made by policy-makers and other agents in the private and public domains. In the domains studied directly in the project, these impacts will include improved management of the UK’s oil and gas resources and better protection of citizens and organisations against organised crime and terrorist activity.

Aston has had a leading role in the development of machine learning techniques for data visualisation for over 15 years. These techniques have been embodied in the open-source Netlab toolbox, which has now been downloaded more than 40,000 times and is used for teaching purposes in at least four universities around the world. A data visualisation tool, Data Visualisation and Modelling System (DVMS), is also freely available online.

Posted in News Tagged with: ,

Leave a Reply

Your email address will not be published. Required fields are marked *

*