DATA RELEVANCE: CONCEPTUALIZATION AND THEORIZATION FROM THE PERSPECTIVE OF KNOWLEDGE ORGANIZATION

Abstract

This study explores relevance from a data perspective in the context of knowledge organization. Via conceptualizing relevance in knowledge organization, examining relationships in knowledge organization, and theorizing data relevance, this study breaks the dominance of relevance studies from an information retrieval perspective, fills the theoretical gap of data relevance, and supplements the relevance functioning factors. Nine definitions of relevance in knowledge organization were proposed based on Saracevic’s relevance definition framework after examining knowledge organization activities, including subject analysis, indexing, cataloging, and social tagging. These definitions provided a deconstructive understanding of relevance in knowledge organization and laid the groundwork for relationship studies between data, and relationship studies between data and information professionals. Relationships in knowledge organization were examined based on the meta-typology of relationships, including relationships in the bibliographic universe, the information universe, and the data universe. Relationships in the bibliographic and information universes showed a trend toward transitioning to data relationships. In the data universe, the concept of relationships has been generalized due to the RDF model; thus, the relationships have been expanded largely with the support of semantic technologies. Relationships explain how data are related and provide a foundation for theorizing data relevance. The data relevance model was created by exploring how information professionals considered data relevance. The user study included a two-stage task and an interview. The grounded theory based data analysis showed 28 functioning factors of data relevance in three levels. The six top-level factors were features of resource, purpose, users, context, potential use, and prior knowledge. The second-level factors showed more data features, such as enabling linkages, scope of terms, granularity, using single/as few schemas, etc. Data relevance theory was developed in the end based on the relevance definitions in knowledge organization, the relationships, and the data relevance model.