RUBRIC is an expert system for full-text information retrieval. The underlying model for RUBRIC's information retrieval process is based upon fuzzy set theory. The RUBRIC developers have compared RUBRIC to the boolean retrieval model, which it subsumes. This study compares RUBRIC to the Vector Space Model for information retrieval, using RUBRIC's own test collection of thirty news articles from the Reuters News Service and their test search for articles which satisfy the information need to find out about "violent acts of terrorism." Results indicate that the vector space model is comparable to RUBRIC for relevant documents, while RUBRIC performs better at retrieving marginally relevant documents.
/lp/association-for-computing-machinery/comparing-vector-space-retrieval-with-the-rubric-expert-system-2JPWOKyLQb