A document classification and retrieval system for R&D in semiconductor industry – A hybrid approach

摘要

In this paper, a hybrid methodology with a vector space model (VSM) and process-oriented attributes for document management is proposed. The VSM is fine-tuned for classifying documents generated during R&D processes. The document correlation values are computed with the VSM for efficient retrieval. Only documents with high correlation values are presented to meet the specific retrieval purpose, which results in efficient and effective document retrieval. We further design a document classification and retrieval prototype system. The prototype is implemented to facilitate R&D document management in semiconductor industries.