The seminar is intended to introduce to each other researchers working on different aspects and applications of link analysis and visualization in order to strengthen the algorithmic foundations of this rapidly emerging, highly interdisciplinary, field.

Link analysis explores associations among entities of arbitrary type. It is increasingly recognized as a fruitful extension of categorical approaches to data analysis in a fast growing number of application domains. Example applications are the analysis of linkages on the Web (search engines, site maps), network traffic monitoring (Web caching, public transport), data mining (e-commerce, telecommunications services), social network analysis (social structures, policy making), text analysis (coreference, cocitation), decision support (financial markets, logistics), or fraud detection (money laundring, calling cards).

Typical objectives in these applications are the identification of central or bottleneck entities, structural patterns and trends, effective modifications, hidden or missing data, substructures, appropriate levels of aggregation, similiarities among data sets, etc., and visualiziation has proven crucial in assisting humans to comprehend complex relational structures and identify unexpected patterns.

Areas of research that are generally relevant are, of course, algorithmic graph theory and graph drawing. In addition, many of the methods used in link analysis have originally been developed in the context of social networks. Since many data sets are quite large and automated support is a necessity, we invited experts on algorithms and data structures for large graphs (external memory, implicit representation) and (graph) algorithm engineering as well. The list of invited participants, many of which have never met before, reflects these core competencies, both on the theoretical and applied side.