SBIR Phase I: Non-Expert Conceptual Annotation

This Small Business Innovation Research (SBIR) Phase I project explores an approach to document indexing and search, which replaces conventional keywords with "concept terms" from a linguistically and semantically rich ontology. Its direct research objectives will be to determine whether the use of such a rich descriptive language can allow non-experts (i.e., non-archivists/librarians) to effectively annotate media resources (especially non-textual media) with "concept terms" which usefully improve both the precision and recall of subsequent searches. The research will prototype technologies and practices that enable semi-skilled workers (drawn from ethnically and economically diverse neighborhoods) to usefully and easily describe content using a conceptual language. The results of this process will be evaluated as to the accuracy of the annotations and their effectiveness in guiding browsing and search. Enabling non-experts (especially content producers) to improve the quality of searches, that is, to create rich and useful metadata, will make the publication and sharing of content more effective and useful. This could be a useful resource for the research community. In addition, the development of the proffered technology may also mean that language technologies can be deployed by enabling a new segment of the work force to perform the important task of concept annotation. Thus, supporting the creation of new semi-skilled entry-level positions in metadata production may broaden the impact of the new economy.