Host-enhanced chemical indexing in technical databases.
Many files that index engineering, physics, or other technical literature contain references to chemical compounds. Complex chemical formulas in the title or abstract often contain special characters, for example, (,), [,], +, -, degrees, and %. Upper and lower case letters often are also included. A search of these formulas and symbols in the basic index is next to impossible because the terms in this field are usually parsed to all alphanumeric characters without sensitivity to case. It is possible for an online host to use a character-recognition algorithm to scan the title and abstract data for special characters or character strings and place them in a separate index field when such files are loaded. Such an algorithm has been designed by Fachinformationszentrun Energie, Physik, Mathematik GmbH in Karlsruhe, West Germany ( FIZ Karlsruhe), the European service center of STN International, the scientific and technical information network. This algorithm recognizes and analyzes chemical formulas, material descriptions, alloys, and eutectic systems as well as nuclear reactions and dopings that appear in the title, abstract, or other fields. These character strings are converted into a standardized form and placed in a new field (the element terms field) which is supplied by the online host during the loading process. A checklist of allowed terms (symbols and chemical formulas) is used to prevent irrelevant terms from being mistaken for legitimate chemical symbols. For instance, the algorithm can recognize that CPU (central processing unit) is not a legitimate chemical formula. It is easy to demonstrate the utility this additional index supplies.(ABSTRACT TRUNCATED AT 250 WORDS)[1]References
- Host-enhanced chemical indexing in technical databases. Badger, E.W., Siems, C.D. Journal of chemical information and computer sciences. (1989) [Pubmed]
Annotations and hyperlinks in this abstract are from individual authors of WikiGenes or automatically generated by the WikiGenes Data Mining Engine. The abstract is from MEDLINE®/PubMed®, a database of the U.S. National Library of Medicine.About WikiGenesOpen Access LicencePrivacy PolicyTerms of Useapsburg