Discussion of index construction for large text collections highlights a new indexing algorithm designed to create large compressed inverted indexes "in situ." Topics include a computational model, inversion, index compression, merging, experimental test results, effect on retrieval performance, memory restrictions, and dynamic…

Presents a theoretical rationale for employing a form of information retrieval algorithm that helps searchers better navigate through large document collections by continually revising probability estimates for document subsets. The algorithm avoids assumptions of index term independence and predicts relevance more accurately with increasing…

Describes the advantages of a prototype retrieval system that uses statistically based ranked retrieval of records rather than traditional boolean methods, especially for end users. Several new techniques are also discussed including bit mapping, pruning, methods of building inverted files, and types of search engine. (26 references) (EAM)

Presents an approach to automatic matching and linkage of citations between databases which applied 14 coding techniques and measured their effectiveness in terms of efficiency of retrieval of duplicates with minimal false hits, reliability, and performance. (EAM)

Describes the use of a highly parallel array processor for pattern matching operations in a bibliographic retrieval system. The discussion covers the hardware and software features of the processor, the pattern matching algorithm used, and the results of experimental tests of the system. (37 references) (Author/CLB)

Describes and discusses the findings of an experiment that tested the precision of full-text retrieval enhanced by 29 document term weighting algorithms at the lowest, middle, and highest levels of recall, and compared with the precision of full-text retrieval without algorithms and with paragraph retrieval. (29 references) (Author/CLB)

Describes an algorithm for computer generation of hierarchical indexes for document collections. The resulting index, when presented with a graphical interface, provides users with a view of a document collection that permits general browsing and informal search activities via an access method that requires no keyboard entry or prior knowledge of…

Zipfian laws for frequency distributions of word pairs and longer phrases are derived from text sample analysis. From crossing of Zipfian curves, it is deduced that number of multi-word phrases that occur frequently in text is surprisingly small, of same order of magnitude as number of individual word-types. (8 references) (EJS)

Describes the indexing concept of entry terms, and evaluates existing terms in the Medical Subject Headings by comparing them to the results of an automatic indexer. The results of experimental testing of a strategy for creating entry terms using the algorithms of the automatic indexer are presented. (31 references) (CLB)

Describes experiments in Arabic information retrieval using 29 queries against a base of 355 bibliographic records. Using roots and stems as index terms worked better than words, and the root performed as well as or better than the stem at low recall levels and definitely better at high recall levels. (Contains 30 references.) (KRN)