Research Theme 3: Advanced Analytics

The four subthemes managed in the Hosting phase of CeADAR at UCD with the leadership of Professor Padraig Cunningham under Research Theme 3 - Advanced Analytics are:

Research Theme: Causation Challenge

This theme aims to create approaches, tools and techneques to improve the accuracy of the detection, identification and communication of cause and effect scenarios across data streams.  Some particular areas for consideration are:  situations where there are multiple causes; measuring the degree of cause for each effect; determining the correlations that influence cause and effect; explore the impact of time between cause and effect; and, identifying intervening or intermediate factors.

 

Research Theme: Social trending and contextualisation challenge

The purpose of this project is to create techniques, tools, and methods to identify trending scenarios in social networks within a defined/relevant context and linking it to internal data.

One tool developed under this subtheme is:

Social Media Entity Monitoring

A state-of-the-art report about this topic is available here:  SOTA on Social Trending & Contextualisation

 

Research Theme: Continuous analytics

The goal of this theme is to develop methodologies, processes, technologies, tools, and algorithms to analyse continuous streams of data using complex analytic algorithms to report the most accurate and timely results.

 

One tool developed under this subtheme is:

Continuous Clustering

A state-of-the-art report about this topic is available here:  SOTA on Continuous Clustering

 

Another tool developed under this subtheme is:

Forecasting Technology Platform

A state-of-the-art report about this topic is available here:  SOTA on Continuous Analytics for Energy Management

 

Research Theme: Social Identity Fingerprint

In order to address the challenge of accurately determining an individual's identity on social networks, this theme's goal is to create techniques, tools and methods to identify and link an individual social media participant across networks using all available information to create an accurate social fingerprint.  In particular, the goal is to enhance current text-based tools that string match to include other ways to identify an individual's social fingerprint such as their geographic location, images, demographic details, and behavioural data.