In this project we aim at formulating enhancing theoretical foundations for the emerging field of graph mining. Graph mining is the field concerned with extracting interesting patterns and knowledge from graph or network structured data, such as can be found in chemistry, bioinformatics, the world wide web, social networks etc. Recent work has shown that many standard data mining techniques can be extended to structured data and can yield interesting results, but also that when applied to complex real-world data, these standard techniques often become computationally intractable. In this project we aim at providing a better understanding of the complexity of the tasks considered in the field of graph mining, and at proposing techniques to better exploit the properties of the data. To this aim, we will bring together insights from the fields of data mining, graph theory, learning theory and different application fields, and add our own original contributions. Key features of the methodology include the ground-breaking integration of insights from graph theory in data mining and learning approaches, the development of efficient prototype algorithms, and the interdisciplinary collaboration with application domain experts to validate the practical value of the work, This potential impact of this project is significant, as it will be the first systematic study of the theory of graph mining, it will provide foundations on which later research can build further and it will have applications in the many domains with complex data.
Fields of science
Call for proposal
See other projects for this call