Pre-defined Terms in Collaborative Indexing Systems - Why Are They Used and Which Impact Do They Have on the Community's Vocabulary?

Collaborative indexing systems have attracted an increasing amount of attention over the last three years. One fundamental limitation to such a system is the uncontrolled nature of its vocabulary, as this consists of terms users freely choose to index resources. As a result, the vocabulary can be poorly structured, making it difficult to identify significant clusters of users, resources, or terms important for search or exploration features. Pre-defined terms are suggested to reduce this uncontrolled vocabulary; these consolidate term usage if users adopt them. However, this suggestion has not yet been proved. This work therefore focuses on an empirical study of the adoption of pre-defined terms. The model of pre-defined term adoption (MPTA) is formulated by applying technology acceptance and diffusion of innovation theories, as well as constructs of social network analysis. MPTA is then validated by one lab experiment with 172 German students, before being cross- validated by two web experiments with 254 German and 160 English-speaking subjects. Results indicate that pre-defined terms are appropriate for reducing the uncontrolled nature of the vocabulary.