Advanced Analytics & Data Mining

Hot:

UE: Financiación de proyectos de inteligencia :

http://cordis.europa.eu/fp7/ict/content-knowledge/fp7-call8_en.html

(…)

Target outcomes
 

a) Reactive algorithms, infrastructures and methodologies (parallelisation, approximation, online processing, compression) for scaling data intensive techniques (including but not limited to machine learning, inference, statistical analysis) up to extremely large data volumes and real time performance. Implementations must be rigorously tested on extremely large and realistically complex data sets coming from diverse resources contributed by organisations with a clear stake in the solution and a clear path to deploying it if effective.

b) Intelligent integrated systems that directly support decision making and situation awareness by dynamically integrating, correlating, fusing and analysing extremely large volumes of disparate data resources and streams. This includes (but is not restricted to) recognising complex events and patterns that are today difficult or impossible to detect, aggregating and mediating opinions or predictions, offering alternative conceptualisations, guaranteeing timeliness, completeness and correctness, integrating categorical and statistical analyses. Visual Analytics should equally integrate data analysis and visualization. The effectiveness of such solutions will be evaluated against the concrete requirements of relevant professionals and communities and tested on appropriately- sized user groups and extremely large data resources from the respective domains (including, but not limited to, finance, engineering, government, geospace, transport, urban management).

c) Framework and tools for benchmarking and exploring information management diversity and comparing and optimising the performance of non mainstream data management architectures and computing paradigms, novel data structures and algorithms on extremely large volumes of data. While methodological rigour and scientific quality and novelty are the main criteria for success, preference will be given to proposals that address a clearly identified industrial, scientific or societal concern or opportunity and/or bring together hitherto unrelated scientific or software engineering communities.

d) Targeted competition framework speeding up progress towards large scale information management systems of global relevance. The framework will be required to: identify a well justified industrial, scientific or societal objective that cannot be attained with the best performing current information management solutions; define detailed experimental conditions under which quantitative progress towards the objective can be reliably observed; implement a fair testing framework inclusive of data resources realistic in size and nature and capable of supporting large numbers of entrants; broadly advertise the competition; administer several testing rounds and publish the outcome of the competition with an appropriate analysis of performance issues and trends.

e) Community building networks and other initiatives designed to link technology suppliers, integrators and leading user organisations. These actions will disseminate results and best practices and address barriers hindering a wider deployment of research results, work towards establishing or advancing widely recognised standards and benchmarks and increase awareness of the potential of the technologies within broader audiences.