Research Questions:

 

The research is centered on the issues that manifest during creation of large semi-structured information management systems, and in application of the semantic web.

 

  1. Distributed Semi-Structured Information Repositories
    1. Query Performance and Scaling

·         What factors most affect semi-structured query performance?

·         What does scaling mean in a distributed semi-structured query system?  How should scaling be measured?

·         What factors most affect scaling of distributed query?

    1. Necessary and Sufficient Constructs

·         How expressive must models be to cover practical, useful, and sale-able use cases?

·         Which use cases can be adequately covered by fully-structured technologies (e.g. XML Databases)?  Which use cases require the simplest levels of the semantic web technology stack (RDF, RDFS)?  Which use cases require richer modeling (DAML, WebOnt, Description Logic)?

    1. Query Languages

·         What is the appropriate scope of languages for semi-structured query?

·         How expressive do queries need to be to cover practical use cases?

·         What is the relationship between various possible query constructs, performance, and scaling?
 

    1. Locus of distribution

·         How do individuals, communities, institutions & enterprises differ in their assignation of value to information?  Is there an appropriate locus of distribution for various types of information?

·         What choices exist for repositories themselves be described, discovered, and opted in/out of the scope of the query?  What are the performance and scaling implications of each choice?

·         What is the relationship between information produced by agents on behalf of the individual, community, institution, or enterprise about a particular asset, and dissemination or transformation services on that asset?  What techniques can be used to discover such information?  What are the implications of different choices about how to distribute such information across various repositories?

 

  1. Metadata Schema Management and Emergent Schemas

·         How can a semantic web of schemas and metadata be effectively managed?

·         Given a particular schema, what techniques can be used to identify schemas that are likely related to it?

·         Given a particular agent or end-user, what techniques can be used to identify schemas that the user is likely to find valuable?

·         How can schema transformations and relationships be described in query-able form?

·         Can schema mapping techniques be demonstrated and provide end-user value in a real system?

·         Are there techniques or best practices (e.g. schema annotation) that can explicitly help balance convergence and specialization?

  1. User Interfaces to Semi-Structured Stores

·         What user interface paradigms and metaphors are useful to end users in interacting with such systems?  In schema creation?  In schema discovery and use?

·         What techniques can reduce the perceived end-user cost of participation in a collaborative information system?