Identifies types and distributions of errors in text produced by optical character recognition (OCR) and proposes a process using machine learning techniques to recognize and correct errors in OCR texts. Results of experiments indicating that this strategy can reduce human interaction required for error correction are reported. (25 references)…

Reports on the performance of the vector space model in the presence of OCR (optical character recognition) errors in information retrieval. Highlights include precision and recall, a full-text test collection, smart vector representation, impact of weighting parameters, ranking variability, and the effect of relevance feedback. (Author/LRW)

Presents results of survey of procedures to handle the circulation of unbound periodicals in 232 public and academic libraries using four major turnkey automated circulation systems. Methods used are described (use of machine-readable symbol attached to issue) and procedures developed at the Eastern Washington University Library are explained in…