participants

• Andrea
• Paolo
• Valerio
• Vasyl
• Aliona
• Alex C
• Alex P

applications

• clustering aka FOAF or Voronoi partitioning
• proximity/distance from Paper to User (recomm.), Paper to Paper (clustering) UserS to Paper (e.g. reviewers), User to Concept (expert - Sub-project 7), User to User
• diversification of search results

• Automatically build (hierarchical) thesaurus of concepts based on their use in papers. Need a module:

  input:   list of papers

  output: hierarchical list of concepts with topics

  
  

  Hierarchically cluster a collection of papers. determines topics.

  Uses: clustering module and possibly DIM module

   

At least 3 Parts

• DIM (direct interaction method)
  • needs to be compared to simpler methods (e.g. co-occurrence)
• biartite clustering (papers connected citations, concepts connected through ontology and papers contain concepts)
  • need to do experiments with paper-2-paper based on citations AND on concepts
• distance on graph

status