Information Storage and Retrieval最新文献

Goffman's indirect method partitions a group of documents into classes based on the number of attributes each pair of documents have in common. The classes produced depend on the threshold value chosen to generate the classes. The indirect method was used here to classify a set of one hundred and nine medical documents, using the index terms assigned by Index Medicus. The attributes used in the indirect method were defined in three ways. First, only exact matches between index terms were considered as attributes in common, resulting in a very fine partition. Next, MeSH tree structure was used to approximate the relationships between index terms, so that the attributes in common were words within a small semantic distance of each other. This produced a broader partition, but several of the documents related in the exact word match dropped below the threshold. To compensate for this, a third definition was used, to give extra weight to exact match relationships. This produced a reasonable classification with all classes nameable. Graphic representations of the three partitions illustrate the structure of the set of documents.

This paper describes a method for automatically segmenting words into their stems and affixes. The process uses certain statistical properties of a corpus (successor and predecessor letter variety counts) to indicate where words should be divided. Consequently, this process is less reliant on human intervention than are other methods for automated stemming.

The segmentation system is used to construct stem dictionaries for document classification. Information retrieval experiments are then performed using documents and queries so classified. Results show not only that this method is capable of high quality word segmentation, but also that its use in information retrieval produces results that are at least as good as those obtained using the more traditional stemming processes.

The growth in number and size of scholarly journals has had a concomitant effect on secondary information services such as Current Contents (CC). To allow an alternative to scanning contents pages, a Weekly Subject Index (WSI) was made available to CC users in 1972. The WSI permits quick access through title words to articles of particular interest and allows retrospective searching of very recent literature. Because of rapid turn-around requirements the production system for the index is computerized with only minimal manual editing done prior to keying. The automated system was recently redesigned to allow the creation of complex bound terms through the use of several stored dictionaries. These include two stop-word, two variant-preferred, and two word-phrase dictionaries, which among them allow up to four sequential words to be combined into one term. Word order within a bound term can be inverted to create a single posting for phrases which can occur in varying sequences. Thus “changes in blood flow” and “blood flow changes” both appear as “blood flow, changes” in the index, avoiding the need for cross-references and double lookups for the user. Such bound terms have greater precision and information content than single title words and their creation increases the efficiency of the WSI for the user.

Science Citation Index (1965–1970) was used to determine the uncitedness of 222 articles published in the Journal of the American Chemical Society (January–February 1965). It was noted that, on the average 14.7 per cent of these test papers were not cited during any given year. It was further observed that only 1 or 0.45 per cent of the 222 papers was not cited at all, during the entire 6-yr period of study. The data obtained in the present study are compared with those of the earlier investigations.