Objective
In this proposal we will offer a unified view of graph mining systems and recommender systems. We propose a system that monitors the time evolution of graph entities (such as customer-product graphs) and allows for a broad set of knowledge extraction operators. The proposed system will feature advanced collaborative tagging techniques, which we call 'auto-tagging' that aim at automating information extraction about the monitored graph entities from various web sources, such as RSS feeds, blog posts and news events. The system will offer several exhiting innovations such as:
- Incorporation of Ontologies and hierarchies on top of the graph structure
- Automatic expansion of entity information from diverse web sources
- Interactive Exploration of the graph space
- Knowledge extraction algorithms, including clustering and bi-clustering, recommendations, anomaly detection and prediction.
Call for proposal
FP7-PEOPLE-2009-RG
See other projects for this call
Funding Scheme
MC-IRG - International Re-integration Grants (IRG)Coordinator
8803 Rueschlikon
Switzerland