Rough and Fuzzy Sets for Data Mining of a Controlled Vocabulary for Textual Retrieval
Padmini Srinivasan, Donald H. Kraft, Jianhua Chen · Studies in fuzziness and soft computing · 2000
We present an approach to text retrieval, incorporating data mining of a controlled i.e., vocabulary mining, in order to improve retrieval Performance. In gener al, formal queries presented to a retrieval System axe not optimized for retrieval efficiency or effectiveness. Vocabulary mining allows us to transform the query via Operations such as generalization or specialization. We offer a new framework for vocabulary mining, combining rough sets and fuzzy sets, allowing us to use rough set approximations when the documents and queries are described us-ing weighted, i.e., fuzzy, representations. We also explore generalized rough sets, variable precision models, and coordinating multiple vocabulary views. Finally, we present a preliminary analysis of the application of our proposed framework to a modern controlled vocabulary, the Unified Medical Language System. The proposed framework supports the systematic study and application of different vocabulary views within the textual Information retrieval model. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.