A STUDY OF EVALUATING THE VALUE OF SOCIAL TAGS AS INDEXING TERMS
Kwan Yi · 2009
AbstractRegardless of their inherent drawbacks as indexing terms, such as semantic ambiguity and the lack of synonymy and polysemy control, the use of social tags is widespread in information organization, filtering, and discovery through social tagging information systems. There are few empirical studies that have evaluated the value of social tags as indexing terms in an operational and experimental setting. The objective of this study is to assess the indexing value of social tags in a context of an information retrieval model using the Latent Semantic Indexing (LSI) method. Socially tagged resources were classified into ten Dewey Decimal Classification (DDC) main classes. Social tags assigned to the resources were used to represent them in LSI. Similarities between resources were measured, and the aggregated similarities according to the ten DDC main classes were compared. Twenty-four different experiments were attempted for optimal results. The empirical data of this study positively demonstrates the value of social tags as indexing term.