The Frequency of Titles Containing "DOPA-Words" - ACS Publications

The use of keywords in titles for the selection of pertinent documents is a commonly used practice,3ranging from the scanning of tables of contents in...
1 downloads 0 Views 224KB Size
F R E Q U E S C Y OF T I T L E S C O S T A I N I N G "DOPA-WORDS

Frequency of Titles Containing "DOPA-words" in a Complete Collection of Published Documents on DOPA (3,4-Dihydroxyphenylalanine) DONALD A. WINDSOR The Norwich Pharmacal C o . P. 0. Box 19 1 , Norwich, N. Y. 13815

Received July 30, 197 1

Seventy-one per cent of the documents in a complete collection of publications on D O P A and dopamine in mammalian systems had " D O P A - w o r d s " i n the titles.

T h e use o f keywords in titles for t h e selection of pertinent d o c u m e n t s is a commonly used practice;' ranging from t h e scanning of tables of contents in primary journals directly or in Current Contents5 t o t h e reading of titles under index terms, a s in I n d e x Mridicus,y t o t h e use of K W I C (key word in context) indexes,4 such as in Biological A b s t r a c t s ' ; or in Chemical Tit 1es. b Resnick found t h a t titles were as effective as abstracts for alerting purposes" a n d Saracevic found t h a t seventy-eight per cent of t h e t i m e titles were just as effective as abstracts or even full texts in conveying relevance t o readers" b u t in only half of Papier's cases did t h e readers assign t h e s a m e title as t h e authors.'" Tell's experiments showed t h a t greater indexing consistency was achieved with titles t h a n with abstracts or full texts.14 Titles as generated by their a u t h o r s are usually not very useful without some additional manipulation. S m i t h increased t h e recall of' titles from Biological Abstracts by using root character strings',' while Flury a n d Henderson obtained successful retrieval with titles only after enriching t h e m . ' After testing Chemical Titles, Chemical-Biological A c t i v i t i e s , a n d A u t o m a t i c S u b j e c t Citation Alert, Abbot, H u n t e r , a n d Simkins did not feel t h a t a pure title search would ever be sufficient.' However, Tocatlian found t h a t t h e titles of chemical papers are becoming more informative.: Garfield h a s wisely cautioned t h a t t h e information content of titles must be judged in t h e context of t h e other citation parameters, s u c h as t h e a u t h o r a n d t h e journal.; To evaluate t h e effectiveness of a keyword in t h e title search it is necessary t o examine a collection already known to contain all of t h e documents on a specific subject. Fortunately. we already have a complete collection of p u b lished papers on D O P A (3,4-dihydroxyphenylalanine),a n d t h e opportunity t o conduct s u c h a n investigation was seized.

MATERIALS For over two years. we have been acquiring all published papers on D O P A a n d d o p a m i n e in m a m m a l i a n systems t o support a research program in t h e use of levodopa ( L - D O P A ) for t h e t r e a t m e n t of parkinsonism.? T h e papers are cited appropriately'" a n d are indexed in d e p t h . At t h e t i m e of this study, there were 1310 documents in t h e collection.

METHODS E a c h citation was examined visually for a t least one of these "DOPA-words": DOPA; Dopamine; 3,4-Dihydroxyphenylalanine; 3,4-Dihydroxyphenethylamine;3-Hydroxyt y r a m i n e ; 3-Hydroxytyrosine; L-DOPA; Levodopa; Laevodopa.

RESULTS Of t h e 1310 papers in t h e collection, 932 h a d a t least one "DOPA-word" i n t h e title for a frequency of 71%.

CONCLUSION O u r collection of published papers on D O P A - d o p a m i n e in m a m m a l i a n systems would be 71% complete if it were constructed only on t h e basis of "DOPA-words" in t h e titles. Increasing t h e keywords t o include s u c h relevant concepts a s "parkinsonism" or "catecholamine" would. undoubtedly have uncovered m a n y of t h e papers in t h e 29% category. However. t h e n u m b e r of false drops would have been m u c h too high t o tolerate. H a d our collection been acquired on t h e basis of a title search alone, most of t h e remaining papers would have been detected eventually from references a n d from secondary journals. Severtheless, when t i m e is i m p o r t a n t a n d it is essential t o know a b o u t a p a p e r as soon as it is published, there is no substitute for i n - d e p t h journal scanning. However, t h e results reported here clearly indicate t h a t more reliance c a n b e m a d e on titles a n d t h e a m o u n t of i n - d e p t h scanning c a n b e greatly reduced.

LITERATURE CITED Abbot. M . T. J . , Hunter, P. S., and Simkins. hl. A., "Current Awareness Searches on CT, CBAC and ASCA," .&lib Proc. 2 0 ( 2 ) ,129-43 (19681. Barbeau. A,, and McDowell. F. H., Ed., "L-DOPA and Parkinsonism." F. A . Davis Co.. 133 pp.. Philadelphia, Pa., 1970. Black, J . D., "The Keyword; Its Use in Abstracting. Indexing and Retrieving Information." Aslib Proc. 14, 313-21 (196.'). Fischer. M.,"The KWIC Index Concept: A Retrospective L'iew," Amer. Doc. 17, 3 - 7 0 (19661. Flury, IF'. R.. and Henderson. D. D.. "A User Oriented Kh'IC Index; KWOC-ed, Tagged and Enriched." Proc. Amer. Soc. Inform. Sei. 7, 101-5 (1970). Freeman, R. R.. and Dyson. G . M . , "Development and Production of Chemicai Titles, a Current Atvareness Index Publication Prepared with the Aid of a Computer." J. Chem. Doc. 3, 16-20 (19631. Garfield. E., "Methods and Objectives in Judging the Information Content of Document Titles." J . Chem. Doc. 10, 260 i1970). Garfield, E.. "Current Contents receives First Information Industry Hall-Of-Fame Award-and Publishes Its 700th Consecutive Issue,'' Current Contents 14(23). 5 (1971). Sational Library of Medicine, "Guide to MEDLARS Services." P. H. S . Publ. T o . 1691, 22 pp., 1967. Journal of Chemcal Documentation. Vol 1 1 . No 4. 1971 227

YAMAMOTO, KUMAI, NAKAKO, ET AL. (10) Papier, L., "Reliability of Scientists in Supplying Titles: Implications for Permutation Indexing," Aslib Proc. 15, 333-7 (1963). (11) Resnick, A., "Relative Effectiveness of Document Titles and Abstracts for Determining Relevance of Documents," Science 134, 1004-6 (1961). (12) Saracevic, T., "Comparative Effects of Titles, Abstracts and Full Texts on Relevance ,Judgments," Proc. Amer. Soc. Inform. Sci. 6, 293-9 (1969). (13) Smith, P. P.. "Increasing Recall of Keyword Indexes by Automatically Indexing Root Character Strings." I b i d . , 7, 275-7 (1970).

(14) Tell, €3. V., "Document Representation and Indexer Consistency: A Study of Indexing from Titles. Abstracts and

Full Text using UDC and Keywords." I b i d . , 6,285-92 (1969). (15) Tocatlian. J. J., "Are Titles of Chemical Papers Becoming More Informative?." J . A m e r . Soc. Inform. Sci. 2115). 345-50 (1970). (16) Rindsor. D. A , , "A Format for the Standardized Citation of Both Published and Unpublished Documents." Ibid . 21(3). 195-7 (1970). (17) Zabriskie, K . H., and Farren. A , , "The B. A . S.I. C. Index to Bioiogicai Abstracts." Amer. J Pharm. E d . 32, 189-200 11968).

Todai Scientific Information Retrieval (TSIR-1) System. I. Genera tion, Updating, and Listing of a Scientific Literature Data Base by Conversational Input T A K E 0 YAMAMOTO.* TOMOKO KUMAI. K E N - I C H I NAKANO. CHlKATAMl IKEDA, TOSIYASU L K U N l l HlDETOSl TAKAHASI. and SHIZUO FUJIWARA

The University of Tokyo, Hongo, Tokyo, J a p a n Received August 30, 197 1

A file structure for scientific literature data, STF, to be used in a scientific information retrieval system, TSIR-1, was developed by modifying the C A S Standard Distribution Format (SDF). The use of STF allows one to merge data records originating in CAS and those generated locally. A set of programs for the generation, updating, and listing of a scientific literature data base in STF from a TSS terminal in conversational mode is described.

For a scientist who wants t o accumulate a d a t a base of scientific literature for his own use a n d for t h e use of his scientific c o m m u n i t y , t h e CAS SDF'. 2 , file structure has several attractive characteristics. It is a highly flexible format a n d is largely based on t h e natural language in its expression of t h e content of t h e literature. However, it still leaves m u c h t o be desired for generating new d a t a records or for d a t a records obtained from other sources t h a n CAS as i n p u t t o t h e d a t a base. We have been building a scientific information retrieval system ( T S I R - 1 ) using a HITAC 5020 T S S 4 ,', of t h e Computer Centre, University of Tokyo. T h e T S I R - 1 system is based on a d a t a structure named STF (for S i m plified Todai F o r m a t , Todai being t h e abbreviated J a p a nese word for t h e University of Tokyo). S T F is closely related t o SDF,a n d its logical content is similar t o t h a t of SDF except for a n u m b e r of modifications. Although more detailed description of STF will b e published elsewhere, significant modifications follow.

The text of the abstract and the full text of an article are given their own identification (ID) numbers and may be recorded in STF files. Both Japanese (Romanized) and English entries are allowed for natural language input. The search programs and queries are expected to take care of the occurrence of the two languages in the retrieval phase. The SERIAL KUMBER data element,'ID number 0012 01, and the Temporary Abstract Sumber (TAS) data element, ID number 0054 01, in SDF' are modified to allow the use of these data elements for the file key in STF. The modification allows one to distinguish, if necessary, between the locally generated data records and records originating. in CAS SDF files, after they have been merged into a unified data base. Thus, the compatibility as well as the distinguishability of the two kinds of records are obtained at the same time. Only capital letters are used in the alphabet. This restriction was necessary because our input-output facilities do not include the lower-case alphabet

In STF, all keywords corresponding to an article item are combined into one record element, and the element is included in the same logical record as the other data related to the item. Thus, each logical record corresponds t o one article except for the case when the data are too great for one logical record (maximum record length: 3520 characters).

In t h e rest of t h e present paper, a set of programs for t h e generation, updating, a n d listing of STF d a t a as TSS disc files by conversational i n p u t will be described. T h e programs were written t o help T S S users t o generate a n d maint a i n their own literature files, without having a detailed knowledge of SDF or of S T F . T h e programs were written in PL/IW language,', a s u b set of PL/I. They were compiled a n d executed by t h e HITAC 5020 TSS.

'To whom correspondence should be addressed a t Department of Chemistry, Faculty of Science. t h e University of Tokyo. Hongo. Tokyo 113. Japan.

228 Journal of Chemical Documentation. Vol. 1 1, No. 4 , 197 1