Data Mining & Social Computing

Eric Robson
Research Unit Manager
Email: erobson@tssg.org

The Data Mining and Social Computing Research Unit focuses on accurately discovering the maximum amount of knowledge in the most efficient manner from increasingly large databases. To achieve this we research the full data analysis life cycle – from data gathering to data cleaning and warehousing through to modelling and presentation of results.

This study is combined with research in Social Computing where we investigate methods for modelling social behaviour through computational systems. The team uses machine learning techniques combined with graph and matrix theory to build representations of social networks. These representations are further mined to give an insight of how people interact with existing ICT services and infrastructure in order to predict and shape the communication networks and services of the Future Internet.

Recent successes in this area are in the launch of Spin-Out Company ZolkC which uses pervasive technology designed and developed in TSSG and the licensing of social network analysis software to Data Fusion International.

The Data Mining and Social Computing Research Unit specialises in techniques such as:

Social Networking Analysis: network and matrix theory.
Predictive Analysis: Classification, Clustering & Association rules mining.
Data Warehousing: ETL, star and snowflake schemas.
Semantic Analysis: Knowledge representations, Annotation, Meaning Discovery, Ontology matching.
Context Awareness: Information & Data Modelling, Location Based, Smart Spaces, Context Federation & Brokerage.
Preferences & Personalisation: Dynamic and static Preference modelling.