CISI.ALL

1255	1	44
.I 45
.T
Measuring the Reliability of Subject Classification by Men and Machines
.A
Borko, Harold
.W
   Procedures for measuring the consistency of document classification are 
described.. Three subject specialists classified 997 abstracts of psychological 
reports into one of eleven categories.. These abstracts were also mechanically
classified by a computer program using a factor score computational procedure.. 
Each abstract was scored for all categories and assigned to the one with the
highest score.. The three manual classifications were compared with each other
and with the mechanical classifications, and a series of contingency 
coefficients was computed.. The average reliability of manual classification 
procedures was equal to .870.. The correlation between automatic and manual 
classification was .766..
.X
26	1	45
28	1	45
45	6	45
61	1	45
73	1	45
81	1	45
82	1	45
174	1	45
175	1	45
310	1	45
315	1	45
419	1	45
420	1	45
421	1	45
422	1	45
454	1	45
479	1	45
483	1	45
485	1	45
486	1	45
487	1	45
562	1	45
564	2	45
661	1	45
662	3	45
663	3	45
664	1	45
769	1	45
853	1	45
1279	1	45
1327	1	45
1419	1	45
1419	1	45
.I 46
.T
A Quantitative Measure of User Circulation Requirements and Its Possible
Effect on Stack Thinning and Multiple Copy Determination
.A
Trueswell, Richard W.
.W
   The strategy outlined in this article describes a possible aid to the 
librarian for thinning a library's stack according to the criterion of user 
needs.. The method uses the last circulation date as a parameter of user 
circulation requirements.. The resulting stack collection would, by design,
satisfy over 99 percent of the user circulation requirements and yet be of
minimum size.. Preliminary, but not yet fully validated research, indicates 
that the number of volumes in a library's holdings may be reduced by 60 to 70 
percent and yet satisfy well over 99 percent of the user requirements.. It may
also be possible to determine which books should have multiple copies in order
that user disappointment can be minimized.. The effective result of this 
approach is that there now may be a quantitative method of maintaining the 
library's holdings at a reasonable level and, in addition, a reduction in the 
disappointment of the user who is unable to find the book he wants because of 
use by another..
.X
31	3	46
33	1	46
36	2	46
41	1	46
46	14	46
115	1	46
181	1	46
182	1	46
183	1	46
184	1	46
193	1	46
195	1	46
198	1	46
201	1	46
206	1	46
207	1	46
208	1	46
266	1	46
269	3	46
278	1	46
280	2	46
282	1	46
290	2	46
367	1	46
395	1	46
415	1	46
759	1	46
760	1	46
767	2	46
774	1	46
778	2	46
891	1	46
905	1	46
925	2	46
943	1	46
952	1	46
953	1	46
964	2	46
968	1	46
977	1	46
983	3	46
1009	1	46
1018	1	46
1019	4	46
1023	1	46
1030	1	46
1086	1	46
1203	2	46
1240	1	46
1359	1	46
1397	1	46
1416	2	46
1417	3	46
1450	1	46
1450	1	46
.I 47
.T
Improvement of the Selectivity of Citation Indexes to Science Literature
Through Inclusion of Citation Relationship Indicators
.A
Lipetz, Ben-Ami
.W
   Citation indexes to large bodies of science literature can often list far
more citing references under the known cited reference than the user can 
afford the time to look up.. By providing some additional information, beyond 
the minimal association of the citing reference with the cited reference, a 
citation index could provide the means of the user to select from a long list 
those citing references which are most relevant to his immediate search 
requirement.. Means of providing this selectivity are discussed.. Particular 
attention is given to the possibility of adding short codes to the citation 
entries which would be informative of the way in which the citing publication
is operationally related to the cited one (this method is an integral feature 
of the Shepard's legal citations).. A scheme of citation relationships of 
potential value to users of science literature is presented.. These 
relationships were tested on a sampling of physics literature.. The suggested 
categories include indicators of the relation of the citing reference to the
scientific process in general, as well as indicators of its relationship to the
cited reference in particular.. Assignment of the categories to a citation 
requires the exercise of judgment, as in subject indexing, but does not involve
the use of subject terminology.. An illustration is provided of the application 
from a citation index to physics literature..
.X
19	1	47
37	1	47
39	1	47
40	1	47
41	1	47
43	1	47
47	6	47
48	1	47
50	1	47
65	1	47
75	1	47
88	1	47
97	1	47
102	2	47
103	1	47
106	1	47
113	1	47
233	1	47
253	1	47
313	1	47
359	1	47
377	2	47
379	1	47
395	1	47
503	1	47
505	1	47
513	1	47
560	1	47
573	1	47
586	1	47
618	1	47
632	2	47
635	1	47
667	1	47
748	1	47
749	1	47
751	1	47
764	2	47
765	1	47
777	1	47
778	2	47
782	1	47
791	1	47
804	1	47
805	1	47
893	1	47
952	1	47
958	1	47
1016	1	47
1061	1	47
1085	1	47
1086	1	47
1087	1	47
1182	1	47
1200	1	47
1210	1	47
1274	2	47
1277	2	47
1278	2	47
1280	2	47
1285	1	47
1287	1	47
1300	1	47
1301	2	47
1302	2	47
1304	1	47
1308	1	47
1313	3	47
1338	1	47
1344	1	47
1347	1	47
1380	1	47
1428	1	47
1444	2	47
1444	2	47
.I 48
.T
The Norms of Citation Behavior: Prolegomena to the Footnote
.A
Kaplan, Norman
.W
   The publication explosion has focused renewed attention on the lowly 
footnote.. While we are all at least partially aware of the technical functions
of the citation for both the writer and the reader of the scientific paper, 
little is known about the norms operating in actual practice.. Even less is 
known about the operating norms with respect to the more sociological functions,
including the acknowledgement of intellectual debts or the conferral of 
recognition upon the works of others.. This paper reviews the state of our 
knowledge, raises questions, and proposes suggestions and hypotheses for 
studying the relationships between footnoting practices among scientists and
the social system of science..
.X
33	2	48
47	1	48
48	10	48
98	2	48
100	1	48
102	3	48
104	2	48
106	3	48
108	1	48
110	1	48
111	1	48
112	1	48
113	4	48
167	1	48
170	1	48
233	1	48
314	1	48
361	1	48
439	1	48
440	1	48
456	3	48
503	1	48
513	1	48
533	2	48
545	2	48
560	1	48
614	1	48
635	1	48
638	1	48
748	1	48
753	1	48
756	1	48
757	1	48
784	1	48
787	1	48
792	1	48
793	1	48
821	1	48
831	1	48
952	1	48
953	1	48
1062	1	48
1082	1	48
1083	1	48
1087	1	48
1173	1	48
1210	2	48
1254	1	48
1256	1	48
1260	1	48
1270	1	48
1273	1	48
1275	1	48
1278	1	48
1285	3	48
1286	2	48
1287	4	48
1291	1	48
1300	1	48
1301	1	48
1302	1	48
1304	1	48
1308	1	48
1313	1	48
1335	1	48
1337	2	48
1338	4	48
1340	1	48
1341	1	48
1344	2	48
1346	2	48
1347	1	48
1369	1	48
1373	1	48
1386	1	48
1418	1	48
1418	1	48
.I 49
.T
Adaptive Information Dissemination
.A
Sage, C.R.
.A
Anderson, R.R.
.A
Fitzwater, D.R.
.W
   Computer dissemination of information offers significant advantages over
manual dissemination because the computer can use strategies that are 
impractical and in some cases impossible for a human.. This paper describes
the Ames Laboratory Selective Dissemination of Information system with emphasis
on the effectiveness of user feedback.. The system will accept any document, 
abstract, keyword, etc., in a KWIC or Science Citation Index Source format..
User profiles consist of words or word clusters each with an initially 
assigned significance value.. These values are used in making the decision to
notify a user that he may be interested in a particular document.. According to
responses, the significance values are increased or decreased and quickly 
attain an equilibrium which accurately describes the user's interests.. The
system is economical compared to other existing SDI systems and human
intervention is negligible except for adding and deleting profile entries..
.X
18	1	49
34	1	49
49	5	49
53	1	49
54	1	49
59	1	49
164	1	49
202	1	49
213	1	49
224	1	49
243	1	49
421	2	49
465	3	49
466	1	49
480	1	49
490	1	49
491	1	49
506	1	49
507	1	49
510	1	49
512	1	49
591	1	49
595	1	49
603	1	49
604	1	49
622	1	49
623	1	49
629	1	49
633	1	49
639	1	49
659	1	49
676	2	49
711	1	49
714	1	49
722	1	49
723	1	49
726	1	49
728	1	49
730	1	49
731	1	49
732	1	49
748	1	49
809	1	49
810	1	49
813	1	49
814	1	49
820	1	49
822	1	49
828	1	49
870	1	49
879	1	49
1091	1	49
1279	1	49
1283	1	49
1298	1	49
1299	1	49
1363	1	49
1366	1	49
1367	1	49
1368	1	49
1396	1	49
1396	1	49
.I 50
.T
Comparison of the Results of Bibliographic Coupling and 
Analytic Subject Indexing
.A
Kessler, M.M.
.W
  A detailed comparison of how 334 papers in Volume 112 of Physical Review form
related groups according to two criteria of relatedness.. The criteria are: (1)
the Analytic Subject Index as used by the editors of the Physical Review; and
(2) the method of Bibliographic Coupling.. The similarities and differences 
between the groups thus formed are illustrated and discussed..
.X
39	2	50
41	1	50
43	1	50
47	1	50
50	8	50
65	1	50
75	1	50
105	1	50
131	2	50
175	2	50
233	1	50
326	1	50
382	1	50
395	1	50
446	1	50
473	3	50
485	2	50
486	1	50
503	1	50
572	1	50
586	1	50
616	1	50
632	1	50
769	1	50
773	1	50
778	1	50
780	1	50
785	1	50
791	1	50
893	1	50
907	1	50
958	2	50
965	1	50
966	1	50
1123	1	50
1207	2	50
1255	1	50
1277	1	50
1280	1	50
1283	1	50
1285	3	50
1287	2	50
1341	1	50
1426	1	50
1426	1	50
.I 51
.T
An Experiment in Automatic Indexing
.A
Damerau, Fred J.
.W
   This report describes a method of indexing documents which is based on the 
assumptions, (1) that a subset of the words in a document can be an effective 
index to that document, and (2) that this subject can be approximated by 
selecting those words from the document whose frequencies are statistically
unexpectedly high.. The results of the experiment are encouraging, although not
definitive because any index set chosen must be tested by using it for 
retrieval from a large collection..
.X
26	1	51
51	5	51
61	1	51
69	2	51
71	1	51
77	1	51
79	2	51
114	1	51
144	1	51
168	1	51
175	2	51
176	3	51
315	1	51
382	1	51
419	2	51
441	1	51
448	1	51
458	1	51
480	1	51
483	1	51
484	1	51
486	1	51
488	1	51
491	1	51
493	1	51
503	1	51
507	1	51
509	1	51
510	1	51
512	1	51
517	1	51
520	1	51
522	1	51
527	1	51
528	1	51
531	1	51
565	2	51
566	2	51
577	2	51
581	1	51
596	1	51
603	1	51
608	1	51
633	1	51
644	1	51
659	1	51
660	3	51
662	2	51
664	1	51
715	1	51
754	1	51
790	1	51
805	1	51
809	1	51
810	1	51
812	2	51
813	1	51
814	1	51
817	1	51
824	2	51
825	1	51
894	1	51
1051	1	51
1294	2	51
1327	2	51
1419	1	51
1427	1	51
1427	1	51
.I 52
.T
Comparative Indexing: Terms Supplied by Biomedical 
Authors and by Document Titles
.A
Schultz, Claire K.
.A
Schultz, Wallace L.
.A
Orr, Richard H.
.W
   The original aim of this study was to obtain objective data bearing on the 
much argued question of whether author indexing is "good".. Author indexing of
285 documents reporting biomedical research was scored by comparing the author-
supplied terms (author set) for each paper with a criterion set of terms that 
was  established by asking a group of 12 potential users to describe the same
document.. Terms in the document title (title set) were scored similarly.. The
average author set contained almost half of all the terms employed by more than 
one member of the user group and scored 73% of the maximal possible score, as 
compared with 44% for the average title set.. When judged by the method and 
criterion employed here, author indexing is substantially better than indexing 
derived from document titles.. The findings suggest that indicia supplied by an
author should serve scientists in biomedical disciplines other than his own 
about as well as they serve his disciplinary colleagues.. The general method
developed for measuring indexing quality may represent a practical yardstick of 
wide applicability..
.X
35	1	52
38	2	52
42	1	52
43	1	52
52	6	52
61	1	52
70	1	52
73	2	52
75	1	52
81	2	52
82	1	52
95	1	52
150	2	52
420	1	52
429	1	52
455	1	52
510	1	52
578	1	52
582	2	52
589	3	52
594	1	52
595	1	52
603	2	52
655	1	52
656	1	52
657	1	52
722	2	52
781	1	52
799	1	52
830	1	52
893	1	52
1054	1	52
1144	1	52
1154	1	52
1195	2	52
1232	1	52
1255	1	52
1281	1	52
1295	1	52
1298	1	52
1421	1	52
1421	1	52
.I 53
.T
The KWIC Index Concept: A Retrospective View
.A
Fischer, Marguerite
.W
   This paper defines and describes the KWIC (keyword in context) index concept,
providing a history of the concept and of its literature.. It discusses 
variations of the index, such as the Bell Telephone Index, KWOC indexes, and 
the WADEX..
   The paper discusses improvements and variations to the KWIC index, such as 
manipulation of the index line, variations of the code, addition of 
classification information, combination of author index and title index, and 
improvements to the preparation of the KWIC indexes, such as improvement of 
titles and use of a thesaurus, and discusses improvement of the use of the KWIC 
index.. The paper discusses the usage of the KWIC index and comments on the 
future of KWIC indexes of the KWIC concept..
.X
18	1	53
34	4	53
49	1	53
53	7	53
59	1	53
150	1	53
164	1	53
175	1	53
202	1	53
213	1	53
224	1	53
243	1	53
252	1	53
421	1	53
465	1	53
466	1	53
480	1	53
490	1	53
491	1	53
506	1	53
507	1	53
510	1	53
512	1	53
565	1	53
589	1	53
591	1	53
595	1	53
603	1	53
604	1	53
622	1	53
623	1	53
629	1	53
633	1	53
639	1	53
659	1	53
674	1	53
676	1	53
680	1	53
711	4	53
715	2	53
722	1	53
723	1	53
726	1	53
728	1	53
730	1	53
731	1	53
732	1	53
746	1	53
760	1	53
809	1	53
810	1	53
813	1	53
814	1	53
820	1	53
822	1	53
828	1	53
850	1	53
870	1	53
879	1	53
1045	1	53
1091	1	53
1118	1	53
1195	1	53
1281	2	53
1283	1	53
1298	1	53
1299	1	53
1327	1	53
1363	1	53
1366	1	53
1367	1	53
1368	1	53
1396	1	53
1396	1	53
.I 54
.T
Simulation of Boolean Logic Constraints Through
the Use of Term Weights
.A
Brandhorst, W.T.
.W
The evolution described below of one aspect of the NASA
system and Technical Information Facility's machine search
system may be of general interest to the documentation
profession.
.X
49	1	54
54	7	54
73	1	54
150	1	54
319	2	54
465	1	54
480	1	54
509	1	54
519	1	54
565	1	54
625	1	54
660	2	54
714	2	54
748	1	54
773	1	54
780	1	54
785	1	54
810	2	54
1279	1	54
1307	1	54
1427	1	54
1427	1	54
.I 55
.T
Bradford's Law and the Keenan-Atherton Data
.A
Groos, O.V.
.W
Bradford's methods are applied to the Keenan-Atherton
data.  The results do not fit Bradford's Law.
.X
55	5	55
210	1	55
219	1	55
225	1	55
359	1	55
361	1	55
379	1	55
573	1	55
592	1	55
614	1	55
747	1	55
748	2	55
751	2	55
777	1	55
778	2	55
791	1	55
792	1	55
889	1	55
893	1	55
1016	1	55
1085	3	55
1086	2	55
1200	1	55
1285	1	55
1337	1	55
1338	1	55
1418	1	55
1418	1	55
.I 56
.T
Computer-Produced Micro-Library Catalog
.A
Kozumplik, W.A.
Lange, R.T.
.W
  The philosophy, production, and cost-effectiveness of
a computer-generated library catalog is described.
This catalog is unique in that it utilizes direct computer
to microfilm composition techniques, employing the
Stromberg Carlson 4020.  Cost, user acceptance, and
by-product capabilities are stressed.
 
.X
56	7	56
341	1	56
886	2	56
887	1	56
892	3	56
1229	2	56
1269	1	56
1371	2	56
1371	2	56
.I 57
.T
Distribution of Indexing Terms for Maximum Efficiency 
 of Information Transmission
.A
Zunde, Pranas
.A
Slamecka, Vladimir
.W
   A function was developed for the optimum distribution of indexing terms by 
the number of postings.. This makes it possible to transmit information with 
maximum efficiency.. The comparison of the actual distribution of the term 
groups with the calculated optimum distribution provides an objective measure 
for evaluating any indexing system with respect to its efficiency as 
information transmission channel..
.X
36	1	57
44	3	57
57	8	57
61	1	57
73	1	57
75	1	57
76	1	57
77	1	57
81	1	57
175	1	57
176	2	57
184	1	57
193	1	57
195	1	57
201	1	57
203	1	57
204	1	57
205	1	57
233	1	57
267	1	57
274	1	57
329	2	57
359	2	57
389	1	57
390	2	57
395	1	57
416	2	57
445	1	57
449	1	57
450	1	57
458	1	57
511	1	57
521	1	57
565	4	57
567	1	57
572	1	57
577	1	57
587	1	57
595	1	57
596	1	57
608	1	57
619	1	57
625	1	57
634	1	57
643	1	57
644	1	57
649	1	57
652	1	57
660	1	57
667	1	57
748	2	57
750	1	57
751	1	57
752	1	57
754	1	57
759	2	57
765	2	57
767	1	57
778	2	57
780	1	57
785	1	57
787	2	57
791	2	57
792	1	57
793	1	57
800	1	57
810	1	57
812	3	57
813	1	57
814	1	57
829	1	57
830	1	57
870	1	57
875	1	57
893	1	57
894	2	57
895	1	57
928	1	57
963	1	57
990	1	57
991	1	57
1081	2	57
1082	2	57
1083	2	57
1085	1	57
1086	1	57
1154	1	57
1201	2	57
1255	2	57
1278	1	57
1282	1	57
1401	1	57
1413	1	57
1417	1	57
1418	1	57
1445	1	57
1445	1	57
.I 58
.T
Relevance Disagreements and Unclear Request Forms
.A
O'Connor, John
.W
   Disagreements about the relevance of documents to retrieval requests occur
because relevance judges differently interpret requests of documents.. Requests
may be differently interpreted because they are unclear.. Well-known types of
request obscurity are reviewed.. Less well known is that a request may be 
unclear because its form - "documents about subject S", "document answering 
question Q", etc. - is unclear.. 
   Explications are developed of the meanings of the request forms just given 
and several others.. A request of any of the forms discussed is interpreted to 
be for documents which support statements of a specified kind in a specified 
way.. Examples are given which suggest that some, perhaps all, "about S" 
requests are unclear.. Some ways of formulating clear question requests are 
given..
   Various ways in which documents may support statements are distinguished..
These depend on on such factors as parts of a document use, inference strength,
and background knowledge permitted.. Some possible clear support specifications
are indicated..
.X
29	3	58
35	1	58
42	1	58
43	1	58
58	10	58
66	1	58
68	4	58
70	1	58
84	3	58
117	1	58
165	1	58
172	1	58
274	2	58
386	1	58
444	1	58
445	2	58
447	1	58
449	1	58
451	3	58
457	1	58
467	1	58
474	1	58
486	1	58
532	1	58
578	1	58
590	2	58
604	1	58
609	1	58
625	1	58
656	1	58
660	1	58
683	1	58
689	2	58
762	2	58
764	1	58
785	1	58
814	1	58
893	1	58
901	1	58
967	1	58
1016	1	58
1030	1	58
1045	1	58
1054	1	58
1077	1	58
1084	1	58
1094	1	58
1195	1	58
1201	1	58
1231	1	58
1235	1	58
1281	1	58
1285	1	58
1285	1	58
.I 59
.T
Opinion Paper
.A