130 likes | 221 Views
The CODE Project envisions a future where research thrives on interconnected data, facilitating knowledge transfer and enhancing research capabilities. By creating a reference manager that extracts and enriches scientific knowledge for easy analysis, it aims to revolutionize data economy. Challenges include data extraction quality, entity disambiguation, and user engagement for monetization. The project collaborates with semantic crowdsourcing and aims to federate, aggregate, and share data efficiently. Join us in building a research ecosystem empowered by Linked Open Data.
E N D
Sebastian Bayerl Kai Schlegel MDPS Workshop-8, 11-15 June 2012
CommerciallyempoweredLinked Open Data Ecosystems in Research Outline • CODE Project • Vision • Project Partners • Our Contribution
CommerciallyempoweredLinked Open Data Ecosystems in Research CODE Project Vision in a Nutshell Project Partners Our Contribution supported by the European Commission under the Seventh Framework Program (FP7)
CODE: Vision in a Nutshell “ ” - Isaac Newton Nanigigantumhumerisinsidentes Standing on the shoulders of giants Research builds on the past We pass knowledge, to create new knowledge Lying under a pile of text documents Unconnected data Contradicting facts Missing / hard to find information Can we do better?
CODE: Vision in a Nutshell (2) Yes,we can … • “A reference manager that does your research for you” * • Given textually encoded scientific knowledge • Extract facts • Enrich & combine with existing knowledge • Make it available for further (visual) analysis • Bootstrapping data economy Vision of the CODE framework * overstated
Vision: Use Case 2. Querying LoD & disambiguation suggestions Further example:Gather structures of research papers (e.g. Images, Tables) 3. Present research results 1. User marks an entity
CODE: Challenges But … how ?!? Challenges Algorithmic quality in extraction Entity disambiguation Efficient Linked Open Data Querying and Aggregation Data Warehousing Approaches User Engagement (Marketplace) Motivation High Quality research Monetary turnover Start simple Focus on concrete use-cases
CommerciallyempoweredLinked Open Data Ecosystems in Research Project Partners 1.700.000 users 1.300.000 users Semantic crowdsourcing power !!! Project duration: 1 May 2012 – 30 April 2014
CODE: ArchitecturalOverview Our contribution
CODE: Projected Goals Triplify, Federate, Aggregate, Share
CODE: Prototype Use Case „CLEF data“ Small steps… • Conference and Labs of Evaluation Forum • Goal: Compare existing systems • PAN: Plagiarism Detection Labs editions • Task: Plagiarism detection • Homogeneous data over years • Numeric facts(score, precision, recall,…)
CODE: Research Objectives ToDo List until 2014 • Middleware based retrieval architecture • Repository discovery • Efficient federation of SPARQL • Aggregation and Data Warehousing • LoD Caching • Provenance
CommerciallyempoweredLinked Open Data Ecosystems in Research “ It is better to take many small steps in the right direction than to make a great leap forward only to stumble backward. ” - Old Chinese quote Thank you! Any questions?