Using semantic components to represent and search domain-specific documents An evaluation of indexing accuracy and consistency

We developed the semantic component (SC) model that supplements existing representations of document content by exploiting domain-specific characteristics of document types and content. The model allows users to express queries against domain-specific components of documents as well as against whole documents. We performed a comparative indexing study as an experimental case study using a national health portal as a framework in order to assess the feasibility of SC indexing. Feasibility includes determining indexing time, perceived difficulty of the indexing task, and indexing accuracy and consistency. The findings show that the feasibility of SC indexing appears to be in the same general range as for assigned indexing by use of indexing terms.