1The area of Information Science gathers a lot of important titles, making the area one of the rare ones to work directly with social contexts, humanities and technology.
2In Brazil, there is an official process of classification of national and international journals of great credibility developed by Capes. The publications which are in the list serve as models to be followed on the national level. However, there are authors who disagree with the titles included in that list.
3Hence, we want to investigate why some international titles included in the Web of Science, which are extremely important for the area of Information Science, are not related in Qualis/Capes classificatory list.
4The Qualis/Capes list is annually organized by national committees of researchers of Graduate Programs Stricto Sensu (one per area evaluated) who are recommended by Capes. Such evaluation considers the journals titles used by the authors of that area (Silva, 2009a).
5The result of that evaluation is the qualification of the journal in A (1 and 2), B (1, 2, 3, 4 and 5) or C depending on the weight given to it, i.e., to the same journal is given different weights in the system, according to the thematic area in which the article is published. However, until August 2012 this process was not based on international standards (Silva, 2009b).
6In this sense, such classificatory system presents some failures. For example, the Journal of Documentation, an influential title in Information Science, is not included in the list of Applied Social Sciences I, to which Information Science is related, but placed in the lists of interdisciplinary (B1) and Architecture and Urbanism (B1) areas.
7Regarding such an inconsistency, is it possible to consider that the journals related to the area of Information Science are, in a broad and coherent way, the best international journals in the area?
8Concerning the area of Information Science, the periodicals categorization has some aspects that are difficult to understand, for example, the permanence of journals classified as international which are not indexed in any international base or directory. Also, there are journals that are indexed in repertories and are not included in the list because the Qualis/Capes list is based on the previous year’s production of Graduate Program professors.
9In the first classificatory version of Qualis/Capes, the initial criteria used to qualify journals were the following:
-
It has to be registered in the ISSN ;
-
The publication rules (particularly, the procedures regarding the submission of texts to a group of referees) have to be clear ;
-
It should have an editorial group composed of national and international researchers ;
-
It should maintain periodicity and regularity ;
-
It should have existed for at least five years ;
-
And should also have a group of national and international collaborators (authors) equally distributed (this criterion, however, must differ foreign authors’ translated texts from foreign authors’ original texts) (Capes, 2005).
10The current version of the system requires, besides the items mentioned above, the impact factor of journals, as it is applied for Chemistry, Physics and Biology, and other areas that categorize their publications in journals ISI/JCR and SciELO (Andrade and Galembeck, 2009). Those criteria must be combined with their periodicals indexation in databases and informative resources (Goldstein Junior, 1999), assuring their quality.
11In this sense, it is important to remember what is understood as scientific communication. It can be defined as the exchange of information among scientific community members, including, according to Garvey (1979), activities related to the production, dissemination and use of information, from the moment the scientist conceives an idea for research to the time his research results are accepted as constitutive of scientific knowledge. Through scientific communication the researchers who participate in the communities are kept informed about the area trends, the developed studies and their results.
12The scientific journals play an important role in the communication among researchers. The number of publications of a group can be considered as an indicator of its scientific production. The number of times the papers of a group are referred in publications is the measure of the impact or visibility of those papers. In the journal global system it is possible to distinguish between the “central” international ones and the peripheral ones according to regional interests. The Institute for Scientific Information, through its indexed bases in Web of Science, aims at encompassing most of the important international periodicals.
13One of the functions of the Web of Science is the analysis of scientific production:
Science needs to be considered as a broad social system in which one of the activities is the dissemination of knowledge. Its second activity is assuring the preservation of patterns, and the third one is giving credit and acknowledgement to those whose papers have contributed to the development of ideas in different fields (Macias-Chapula, 1998, 136).
14Among other means of generation of productivity indexes, the best known is bibliometry, which is based on qualitative judgments, reusing evaluations undertaken for other purposes and using quantitative techniques for its treatment.
15The indicators of Input and Output (product), their relationship with each other, as well as their importance for quantitative studies, are discussed by authors such as Spinak (1998). By using those indicators, the scientific activity can be compared to companies’ activities, with inputs and results. According to Población and Oliveira (2006, 68), “input is a combination of factors that favor the production of a determined amount of goods and services (output)”. The measuring becomes the base for production of scientific indicators.
16Regarding the aforementioned considerations about scientific communication, it is worth mentioning that the quality of databases, which are understood as electronic information sources, can be looked up in an interactive or conversational way through electronic resources. Bibliographic sources are characterized by the presence of an organized collection of bibliographic information; machine legibility and possibility to be accessed online or in batch (Cunha, 1997).
17Databases play an important role in the dissemination of information, storing and providing its retrieval. In this aspect they greatly contribute to any investigative process.
18In a research process, scientific visibility is a considerable factor. In the databases’ universe, such factor is also analyzed as information management process and competitive intelligence, precisely because it is inside information.
19There are several databases which work as information sources for specific areas such as: Chemical Abstract for Chemistry, the ERIC for Education, the EconLit for Economics and the LISA and Library Literature for Library and Documentation Sciences. However, a lot of scientific areas use multidisciplinary sources despite the availability of specialized bases, such as Web of Science and more recently, Scopus.
20Web of Science holds three bases: the Science Citation Index Expanded, published since 1964, the Social Science Citation Index, edited since 1972 and the Arts & Humanities Citation Index, organized since 1975. Its repertoire is included in more than 10,000 scientific journals of all fields of knowledge. Web of Science selects high impact’s journals. It is most used for bibliometric studies. Its structure presents information about author(s), institution(s), article title, journal, volume, number, pages, publication year, descriptors and country, among others. It also favors data extraction for elaborating indicators and variables for study. However, this base has some access restrictions. An online interface of the Web of Science is available in Capes Portal but only people who are in Capes system are allowed to use it together with the private universities that pay for its access.
21Given the high impact of information science journals indexed in the Web of Science, we wonder why most of those titles are still not included in the Qualis/Capes list.
22Some journals indexed in the Web of Science are qualified as journals layer A1 to Qualis / Capes, yet they are referenced with this qualification only if a permanent teacher in a graduate program has published in a journal that is indexed in the database.
23Even though the Web of Science is a multidisciplinary database it is an important resource for the field of Information Science since the most relevant contents of the area indexed by LISA, LISTA, ISTA are also indexed in the WoS, (Lucas, Pinto & Lara, 2013).
24Another critical fact is that the list of Qualis / Capes is evolving very slowly with little changes since the years 2000.
25Just for comparison, in Spain, journal´s criteria are structured in three points, where two points are very similar to the Brazilian model:
-
The dissemination of publication and access the teacher’s job for usability overall,
-
The publication journals are peer review, and
-
The difference is in using citation indices and impact as differential (DICE, 2014).
26The Spanish system uses a large number of information sources - Francis, International Bibliography of Social Sciences, Arts and Humanities Citation Index, Social Science Citation Index, Bibliography of the History of Arts (RLG), Historical Abstracts of Music Literature, Philosopher’s Index, Répertoire Bibliographique, International Bibliography of Periodical Literature in Humanities and Social Sciences (IBZ) Bibliographie Linguistique / Linguistic Bibliography (BL) ea Library and Information Science Abstracts (AECI, 2008).
27In France, researchers are incited to publish in journals having:
-
An ISSN or e-ISSN number ;
-
A website with domain ;
-
An institutional support to look at the regular dissemination in a five years range ;
-
Editing data (abstract, keywords) ;
-
Editorial counseling ;
-
A variety of publications and authors (domestic and foreign) ;
-
And journals that are indexed in international databases (AERES, 2013).
28Closer to Brazil, the Argentine Center of Scientific and Technological Information (CAICyT) determines that a journal should have all the key requirements of publication, but resources should be highly visible and indexed in SciELO-Argentina, Scopus or the Web of Science (Flores, 2010).
29In order to clear up that situation, it is important to : (I) identify the most relevant journals for Information Science ; (II) investigate the most representative authors in this field ; (III) identify the authors’ citation index ; (IV) develop co-citation maps ; (V) generate convenient scientific indicators, according to theories of Bradford, Lotka and Zipf.
30As a result of that work, based on bibliometric analysis, it is expected to identify and suggest to Capes more appropriate titles in which Brazilian authors can publish their studies/papers. Such list would hold relevant titles for the Information Science area. By analyzing the most representative authors and their citation index, the second contribution would be to identify the level of recognition and visibility of the authors by their peers as their papers are published in the most representative journals in the Information Science area.
31Therefore, it is important to explain why the most important authors and their co-citations are investigated. First, because it is the aim of the Brazilian government, through Capes, to elaborate a list of journals in which the authors “should” publish their studies/papers. Second, because we want to see to what extent the Web of Science’s content can give us a global overview of the Information Science area and therefore could serve as a guide for Capes to establish a list of journals for authors in Information Science.
32From 1994 to 2008, we have selected some of the highest impact factor journals concerning the area of Library and Information Sciences (Library Science & Information Science) in the world, based on the analysis made by Leydesdorff (2007) and complemented by a research for titles formerly carried out by the authors of this study.
33Our work is about citation and co-citation networks of scientific journals. This characteristic is based on the impact factor of the periodicals of Information Science regarding the relationship and citation index (with) (for) the Scientometrics journal, in which the author describes the most important journals for the area in a comparison to Scientometrics.
34The methodology is developed in two steps: one to explain the search strategy and the other one for informational retrieval and treatment of those data, measured by statistics and bibliometric theories (Vanti, 2002).
35The pattern strategy of data search and retrieval is characterized by a simple research in which the period and all databases of Web of Science (Science Citation Index, Social Science Citation Index e Arts & Humanities Citation Index) are limited, focusing on the field of the journal title (Source Title).
36The retrieved journals are presented by their full name; however, in some cases, the data base filter is used to have the journals simplified names. There is a selection of 28 journals for the analysis purpose.
Table 1: List of ISI journals for Information Science & Library Science areas
Journal
|
Country
|
Language
|
Annual Review of Information Science and Technology
|
USA
|
English
|
Aslib Proceedings
|
ENGLAND
|
English
|
Canadian Journal of Information and Library Science
|
CANADA
|
English
|
College & Research Libraries
|
USA
|
English
|
Government Information Quarterly
|
USA
|
English
|
International Journal of Information Management
|
ENGLAND
|
English
|
Information Processing & Management
|
ENGLAND
|
English
|
Information Society
|
USA
|
English
|
Information Technology & Libraries
|
USA
|
English
|
Journal of Academic Librarianship
|
USA
|
English
|
Journal of the American Society Information Science & Technology
|
USA
|
English
|
Journal of the American Society Information Science (even 2000)
|
USA
|
English
|
Journal of Documentation
|
ENGLAND
|
English
|
Journal of Information Science
|
ENGLAND
|
English
|
Journal of Library & Information Science
|
USA
|
English
|
Knowledge Organization
|
GERMANY
|
English
|
Library Hi Tech
|
ENGLAND
|
English
|
Libri
|
GERMANY
|
English
|
Library & Information Science
|
JAPAN
|
Japanese
|
Library & Information Science Research
|
USA
|
English
|
Library Journal
|
USA
|
English
|
Library Quarterly
|
USA
|
English
|
Library Trends
|
USA
|
English
|
Online
|
USA
|
English
|
Reference User Services Quarterly
|
USA
|
English
|
Scientometrics
|
NETHERLANDS
|
English
|
Zeitschrift fur Bibliothekswesen und Bibliographie
|
GERMANY
|
German
|
37For informational treatment, a strategy developed by Pinto et al. (2007) has been used, in which co-citations networks are mapped. That type of analysis, also known as information visualization (Pinto & Rodríguez-Barquím, 2005), favors data representation.
38For the treatment of retrieved registrations, ProCite is used (Saxton, 2001). In regard to information visualization, CiteSpace (Chen, 2006) is used; it treats and defines indexes automatically.
39For the data visualization, we used a system where the size of the nodes depends upon their centrality. Darker shades are closer to 2008, and clearer are closest to 1994. Another detail, the greater the circumference higher is the citation impact of the unit of analysis.
40This step is centered on the investigation of productivity in the most important journals for areas correlated with Information Science and their visibility in the Web of Science database.
41Among those journals, we notice that only two of them are included in the Qualis/Capes list, Information Processing & Management and Organization Knowledge (list version by local, regional and international level). However, in the list that was released in 10/13/2009, these journals were excluded and the Library Trends journal was added.
42Curiously, there are some initiatives in the country, for example, the School of Communication of the University of São Paulo (ECA/USP) released a book about scientific communities journals and their respective areas, in which Information Science is particularly important because there are two studies almost completed that focus the Qualis/Capes list and national journals, and international journals.
4311,212 authors have published in those journals, representing a total of 19,855 scientific articles from 1994 to 2008.
44Lotka’s Law was used to verify the authors’ productivity and to provide a firm foundation for the initial premise that “few researchers publish a lot of works and a lot of researchers publish few works” to which the inverse Square law was used (1/n²).
45This analysis shows that 20 % of the authors published 10,254 documents, representing 90,9 % of productivity.
46This High index that reveals a production distortion, in which 20 % of scientists produce 80 % of literature, is also known as the inverse square. In the scientific community, it is also known as Pareto’s principle, but its application in Bibliometry might be attached to the elite of authors.
47Another analysis made is the Elitism Law, proposed by Price (1963), which estimates how big a specific population of elite authors is. Is this case, the elite of authors is composed by 105 names, represented on Table 2.
Table 2: Elite of Authors (WoS 1994-2008)
Order
|
Authors
|
Fq
|
Order
|
Authors
|
Fq
|
Order
|
Authors
|
Fq
|
1
|
[Anon]
|
134
|
36
|
Gunter, B
|
22
|
71
|
Moed, Hf
|
15
|
2
|
Tenopir, C
|
104
|
37
|
Wolfram, D
|
22
|
72
|
Rowlands, I
|
15
|
3
|
Nicholas, D
|
85
|
38
|
Chen, Hc
|
21
|
73
|
Vaughan, L
|
15
|
4
|
Oppenheim, C
|
66
|
39
|
Dugan, Re
|
21
|
74
|
Warner, J
|
15
|
5
|
Oder, N
|
61
|
40
|
Mcafee, Rp
|
21
|
75
|
Wilson, Td
|
15
|
6
|
Rousseau, R
|
60
|
41
|
Berry, J
|
20
|
76
|
Davenport, E
|
14
|
7
|
Egghe, L
|
59
|
42
|
Coyle, K
|
20
|
77
|
Moya-Anegon, F
|
14
|
8
|
Thelwall, M
|
58
|
43
|
Large, A
|
20
|
78
|
Fox, Bl
|
14
|
9
|
Huntington, P
|
55
|
44
|
Rowley, J
|
20
|
79
|
Hawkins, Dt
|
14
|
10
|
Rogers, M
|
54
|
45
|
Braun, T
|
19
|
80
|
Hjorland, B
|
14
|
11
|
Hoffert, B
|
52
|
46
|
Delfino, E
|
19
|
81
|
Lewison, G
|
14
|
12
|
Spink, A
|
47
|
47
|
Pearl, N
|
19
|
82
|
Mcknight, C
|
14
|
13
|
Williams, P
|
45
|
48
|
Zhang, J
|
19
|
83
|
Mcmurdo, G
|
14
|
14
|
Tennant, R
|
42
|
49
|
Beheshti, J
|
18
|
84
|
Meyer, M
|
14
|
15
|
Berry, Jn
|
39
|
50
|
Connolly, B
|
18
|
85
|
Minkel, W
|
14
|
16
|
Ojala, M
|
39
|
51
|
Morris, D
|
18
|
86
|
Stlifer, E
|
14
|
17
|
Cronin, B
|
36
|
52
|
Shuler, Ja
|
18
|
87
|
Willett, P
|
14
|
18
|
Crawford, W
|
33
|
53
|
Garg, Kc
|
17
|
88
|
Wilson, Cs
|
14
|
19
|
Leydesdorff, L
|
33
|
54
|
Golderman, G
|
17
|
89
|
Baxter, G
|
13
|
20
|
Glanzel, W
|
30
|
55
|
Ingwersen, P
|
17
|
90
|
Born, K
|
13
|
21
|
Mcclure, Cr
|
29
|
56
|
Nicholson, S
|
17
|
91
|
Burns, A
|
13
|
22
|
Nottelmann, H
|
29
|
57
|
Savolainen, R
|
17
|
92
|
Collins, Br
|
13
|
23
|
Bar-Ilan, J
|
27
|
58
|
Anderson, Aj
|
16
|
93
|
Dimattia, Ss
|
13
|
24
|
Hernon, P
|
27
|
59
|
Burrell, Ql
|
16
|
94
|
Hartley, J
|
13
|
25
|
O’leary, M
|
26
|
60
|
Jansen, Bj
|
16
|
95
|
Klein, Lr
|
13
|
26
|
Albanese, Ar
|
25
|
61
|
Kostoff, Rn
|
16
|
96
|
Makulowich, Js
|
13
|
27
|
Cole, C
|
25
|
62
|
Losee, Rm
|
16
|
97
|
Miller, R
|
13
|
28
|
Dilevko, J
|
25
|
63
|
Schubert, A
|
16
|
98
|
Williams, W
|
13
|
29
|
Bawden, D
|
24
|
64
|
Yang, Cc
|
16
|
99
|
Barry, J
|
12
|
30
|
Bertot, Jc
|
24
|
65
|
Foo, S
|
15
|
100
|
Bates, Me
|
12
|
31
|
Bryant, E
|
24
|
66
|
Jaeger, Pt
|
15
|
101
|
Chowdhury, Gg
|
12
|
32
|
Budd, Jm
|
24
|
67
|
Kenney, Bj
|
15
|
102
|
Gao, Sj
|
12
|
33
|
Ford, N
|
24
|
68
|
Kling, R
|
15
|
103
|
Kumar, S
|
12
|
34
|
Gupta, Bm
|
23
|
69
|
Laguardia, C
|
15
|
104
|
Pack, T
|
12
|
35
|
Jarvelin, K
|
23
|
70
|
Marcella, R
|
15
|
105
|
Perez, E
|
12
|
NOTE: Fq = Frequency; [ANON] = Much are editorial and historical documents of the journal.
48The highest productivity in absolute numbers comes from anonymous authors. This is assumed to happen due to the fact that Web of Science turns some editorial notes into scientific articles automatically. Lots of those papers are written by editors who describe the journals contents as an informative synopsis, which is very common in scientific journals. This is an interesting datum because even if only scientific articles are selected, the editorials will come out since they are considered as articles by the system. That is why some editors appear in the journals publications.
49In the frequency chart of the most important authors, it is also possible to see the centrality of those authors and their collaborators, pointing out that the higher professional visibility the researcher gets, the bigger his/her opportunity to publish new papers (Valchy, 1970).
Figure 1: Centrality and frequency of the most representative authors (WoS 1994-2008)
50On this figure, the size of the circles depends upon the authors’ position in their clusters (groups). The red color shows the centrality among their peers, which is high for Glänzel W, Hernon P, Thelwall M, Davenport E, Williams P, Large A, among others.
51We can distinguish between different thematic areas: the large group of authors who work on Metric Studies; a group involving researchers in Digital Libraries; another group working on discussions about Formation and Profile of Library Science and Information Science; two other groups on Information Retrieval.
52The color shades, from bright to less bright represent the frequency of the author’s publications in the studied journals; from orange shades to yellow ones (2008 to 2003), then, green color, from the brightest to the lightest (2002 to 1999) and blue shades (1998 to 1994). The greater the representation of all color shades, the higher the visibility of the authors throughout the time period.
53The analysis of keywords resulted in 3,209 terms, in which descriptors’ frequency summed up 16,981 appearances.
54According to Zipf’s Law, which measures words frequency, it has been identified a frequency of 50 % in which the totality was represented by 8,654 (a little over 50 %) from 132 keywords considering that all constant data over 20 appearances were included.
55The following information is taken from the fields: ID (identification of descriptors), DE (descriptors) and KW (keywords) in Web of Science.
56Table 3 shows the compound terms (in bold) as the most representative for Information Science area; the single terms are not very representative because they are more generic, such as Work, Future, Needs, among others. Therefore, it is observed that the seeking and indexing system of Web of Science concerning attribution to keywords is not very efficient. More generic indicatives should be considered and attributed automatically. The data consistency and normalization has not been verified.
Table 3: Key-words frequency (WoS 1994-2008)
Thematic area
|
Fq
|
Thematic area
|
Fq
|
Thematic area
|
Fq
|
Science
|
491
|
Impact-Factors
|
52
|
Information-Systems
|
28
|
Information
|
326
|
Scientists
|
51
|
Law
|
28
|
Retrieval
|
294
|
Services
|
50
|
Links
|
28
|
Information-Retrieval
|
244
|
Students
|
50
|
Obsolescence
|
28
|
Systems
|
232
|
Work
|
49
|
Success
|
28
|
Model
|
206
|
Framework
|
47
|
Authorship
|
27
|
Internet
|
199
|
Life
|
47
|
Instruction
|
27
|
Design
|
198
|
Publication
|
47
|
Usage
|
27
|
Library
|
177
|
Education
|
46
|
Information Technology
|
26
|
Behavior
|
169
|
Strategies
|
45
|
Online Catalog
|
26
|
Technology
|
169
|
Algorithms
|
44
|
Ranking
|
26
|
Impact
|
162
|
Productivity
|
44
|
Index
|
25
|
Web
|
157
|
Resources
|
44
|
Space
|
25
|
Knowledge
|
153
|
User
|
43
|
Tasks
|
25
|
Seeking
|
145
|
Issues
|
42
|
Feedback
|
24
|
World-Wide-Web
|
144
|
Publications
|
42
|
Firm
|
24
|
Journals
|
140
|
Research Performance
|
42
|
Project
|
24
|
Indicators
|
138
|
Faculty
|
41
|
Scientific Journals
|
24
|
Communication
|
136
|
Organizations
|
41
|
Women
|
24
|
Performance
|
135
|
Selection
|
41
|
Age
|
23
|
Users
|
133
|
Document-Retrieval
|
40
|
Ask
|
23
|
Relevance
|
126
|
Queries
|
39
|
Bibliographic Instruction
|
23
|
System
|
116
|
Context
|
38
|
Care
|
23
|
Patterns
|
110
|
Interface
|
37
|
Decision-Making
|
23
|
Networks
|
103
|
Academic-Libraries
|
36
|
Electronic Journals
|
23
|
Information-Science
|
93
|
Bibliometrics
|
36
|
Experience
|
23
|
Needs
|
93
|
Perceptions
|
36
|
Output
|
23
|
Management
|
88
|
Trends
|
34
|
Query Expansion
|
23
|
Citation
|
84
|
Co-Citation
|
33
|
Satisfaction
|
23
|
Quality
|
84
|
Countries
|
33
|
Strategy
|
23
|
Citation-Analysis
|
76
|
Sites
|
33
|
Time
|
23
|
Access
|
67
|
Hypertext
|
32
|
Words
|
23
|
Information-Seeking
|
67
|
Scholarly Communication
|
32
|
Attitudes
|
22
|
Classification
|
65
|
Search Process
|
32
|
Computer-Mediated Communication
|
22
|
Libraries
|
63
|
Database
|
31
|
Economics
|
22
|
Models
|
63
|
Organization
|
31
|
Determinants
|
21
|
Search
|
62
|
Environment
|
30
|
Future
|
21
|
Online
|
61
|
Growth
|
30
|
Humanities
|
21
|
Text
|
61
|
Seeking Behavior
|
30
|
Policy
|
21
|
Perspective
|
60
|
World
|
30
|
Scientific Collaboration
|
21
|
University
|
57
|
Documents
|
29
|
Search Engines
|
21
|
Innovation
|
56
|
Humanities Scholars
|
29
|
Service
|
21
|
Collaboration
|
54
|
Articles
|
28
|
Text Retrieval
|
21
|
Data-Bases
|
54
|
Criteria
|
28
|
Yahooligans
|
21
|
NOTE : Fq = Frequency.
57The data have been treated and kept in the original language.
58To analyze the relationship between the keywords, one should be very attentive because, according to Santos and Kobashi (2009, p. 167),
…studies with semantic entities, such as the descriptors of a database, require intense previous treatment of data consolidation before they are submitted to bibliometric processing. In this sense, it is necessary to take a careful look at the automatic data classification because it may not bring much reliable results.
Figure 2: Correlation between keywords and descriptors (WoS 1994-2008)
59Figure 2 shows the keywords’ co-occurrence relationships in the studied journals. For example, the word retrieval alone is very close to the word information, and also close to the compound noun information retrieval.
60Such case can also be applied to some sparse words. By checking the top of the Figure, there are the words journal, citation analysis, productivity, collaboration, indicators and citation, all of them concerning Science Metric Studies.
61In the center of the Figure, there is the library management environment with connected terms such as library, libraries, knowledge, management, needs and quality, among others.
62The results obtained from the keywords’ relationship map confirms what Table 3 shows that a lot of terms seem to have a single application, but actually (compound or not) they help to understand the behavior of Information Science identifiers in the journals studied.
63For the citation analysis, three variables have been verified: the most cited authors, the most cited journals and the most cited documents, showing the frequency (through tables) and the similarity (through co-occurent figures).
64The bibliographic coupling (retrospective) consists in the compilation of articles citing the same document(s), and the co-citation (prospective) in the compilation of articles cited by the same documents. According to Marshakova (1981), the bibliographic coupling measures the coupling level of two or more articles according to the number of identical documents cited by those articles. The co-citation measures the coupling level of two or more articles by the number of documents where those articles are cited, simultaneously.
65Based on that coupling, 27,116 different cited authors are gathering the total of 59,951 appearances.
66Table 4 shows the list of authors who have over 50 cited papers.
Table 4: Index of authors who have over citations (WoS 1994-2008)
Authors
|
Fq
|
Authors
|
Fq
|
SALTON G
|
453
|
KLING R
|
71
|
SPINK A
|
205
|
BLAIR DC
|
70
|
NICHOLAS D
|
190
|
YANG Y
|
69
|
BELKIN NJ
|
189
|
LAWRENCE S
|
68
|
SARACEVIC T
|
161
|
HARTER SP
|
67
|
ROBERTSON SE
|
135
|
TENOPIR C
|
67
|
DERVIN B
|
119
|
MARCHIONINI G
|
66
|
GARFIELD E
|
115
|
NIELSEN J
|
66
|
EGGHE L
|
114
|
PORTER ME
|
65
|
THELWALL M
|
110
|
CRONIN B
|
64
|
HARMAN D
|
105
|
BUCKLEY C
|
63
|
KUHLTHAU CC
|
105
|
FIDEL R
|
62
|
BATES MJ
|
100
|
WILSON TD
|
61
|
DAVENPORT TH
|
96
|
ALLAN J
|
59
|
VANRIJSBERGEN CJ
|
94
|
COOPER WS
|
55
|
JONES KS
|
92
|
NONAKA I
|
55
|
LOSEE RM
|
92
|
ROUSSEAU R
|
54
|
HERNON P
|
90
|
HAWKING D
|
53
|
JANSEN BJ
|
89
|
LANCASTER FW
|
53
|
ELLIS D
|
86
|
EARL MJ
|
50
|
INGWERSEN P
|
84
|
HEARST MA
|
50
|
BORGMAN CL
|
82
|
RICE RE
|
50
|
CROFT WB
|
81
|
SCHAMBER L
|
50
|
VOORHEES EM
|
73
|
|
|
NOTE : Fq = Frequency.
67Considering this volume of citations, it is possible to say that all those authors are extremely important for the area. Not only are they highly cited in the Web of Science journals but also are their publications incorporated in basic bibliographies of Library Science courses in Brazil.
68Moreover, it is essential to highlight their relevance in their thematic areas, for example: Lancaster and Robertson are relevant in indexation and Information Analysis; Rousseau, Garfield, Egghe and Thewmall in Metric Studies; Nonaka I in Informational Management; Cronin, Salton, Spink, Ingwersen and Blair in Information Retrieval, information needs and informational behavior; Nicholas, Belkin and Saracevic in Fundamentation of Information Science; and Dervin in communication theory.
Figure 3: Co-occurrence of authors’ citations (WoS 1994-2008)
69On figure 3, we can notice that there is a very strong relationship between information retrieval (Salton, Ingwersen) and metric studies (Garfield, Small, Price, Moed, Narin and Glänzel). However, there is also a relationship regarding, at the same time, the information retrieval and the metric studies, well represented by Cronin.
70In fact, the network provides explanations about the grounding of Social Science and Information Science.
71Another analysis proposed is that of verifying the citation index of journals by using Bradford’s theory to refine scattered zones. The number of cited journals is 26,014, with a total amount of 63,572 citations. In the first zone, we identified 226 cited journals. For a visual simplification, only the first 45 cited journals are represented, with average citation rates of over 100 appearances.
Table 5: Index of journals citations (WoS 1994-2008)
Cited Journals
|
Fq
|
Cited Journals
|
Fq
|
J AM SOC INFORM SCI
|
1687
|
LIBR TRENDS
|
179
|
INFORM PROCESSING MA
|
1195
|
RQ
|
166
|
COLL RES LIBR
|
782
|
INFORM MANAGE
|
165
|
COMMUN ACM
|
682
|
INFORMATION TECHNOLO
|
156
|
J DOCUEMENTATION
|
642
|
SCIENCE
|
147
|
J ACAD LIBRARIANSHIP
|
420
|
NIST SPEICAL PUBLICA
|
137
|
MIS QUART
|
405
|
ORGAN SCI
|
133
|
J AM SOC INF SCI TEC
|
389
|
J MANAGEMENT INFORMA
|
132
|
SCIENTOMETRICS
|
366
|
LIBR QUART
|
130
|
J INFORMATION SCI
|
338
|
LIBR J
|
127
|
INT J INFORM MANAGEM
|
316
|
INFORM SYST RES
|
125
|
GOV INFORM Q
|
252
|
MIS Q
|
125
|
HARVARD BUS REV
|
251
|
P 21 ANN INT ACM SIG
|
124
|
INFORMATION PROFESSI
|
230
|
ACAD MANAGE REV
|
121
|
INFORMATION RETRIEVA
|
226
|
PUBLIC ADMIN REV
|
116
|
ANNU REV INFORM SCI
|
219
|
RES STRATEGIES
|
116
|
LIBR INFORM SCI RES
|
217
|
BRIT MED J
|
115
|
MANAGE SCI
|
208
|
B MED LIBR ASSOC
|
110
|
ACM T INF SYSTEMS
|
200
|
INFORM SOC
|
110
|
COLL RES LIB
|
197
|
INTRO MODERN INFORMA
|
110
|
ASLIB PROC
|
195
|
WASHINGTON POST FEB
|
108
|
INFORMATION SYSTEMS
|
192
|
COLL RES LIB NEWS
|
100
|
SLOAN MANAGE REV
|
181
|
|
|
NOTE : Fq = Frequency.
72Apart from those 45 most relevant journals, it is important to mention some titles which, although they are not related on the prior Table, belong to Bradford’s first scattered zone, such as Decision Sci; Inform Technol Libr; Online, Information Management; Reference Services R; Lect Notes Comput Sc; Reference Librarian; Acad Manage J; Strategic Manage J, Information Seeking; J Lib Adm; Ann Int Acm Sig; Am Lib; Libr Resour Tech Ser, Lib J; Ann Int Acm Sig; Information Storage; Decis Support Syst; Computational Lingui; Ann Int Acm Sig, Telecommun Policy; Eur J Inform Syst; J Information Techno; J Marketing; Ann Int Acm Sig; Acm Comput Surv; Smart Retrieval Syst; Int J Man Mach Stud; Collection Managemen; J Strategic Inf Syst; Comput J; Internet Res; Res Policy.
73Some of the journals investigated in this study show high citation indexes, which reinforces the fact that journals indexed in the same database tend to be mutually cited. For Instance, the Journal of Information Science, Journal of Documentation, Information Processing & Management and Scientometrics are some of the high impact journals which are often cited together, as it can be seen in Figure 4.
Figure 4: Co-occurrence of journals citation (WoS 1994-2008)
74Focusing on the categorization of scientific centrality in co-citations, we observe that seven journals hold the status of the most influent, and are displayed with a pink outline in Figure 4, Jasist, Communication ACM, Information Processing & Management, Research Policy, College Research Library, RQ and Journal of Documentation.
75The other color shades represent the historical periods of co-citations of those journals; thus, the red and orange shades represent the most recent years (2004 to 2008), the yellow one represents 2003 and 2002, the green shades identify 1999 to 2001 and the several blue shades represent the period from 1994 to 1998.
76Regarding this display, the journals with bigger circles are seen all along since they are essential to the Information Science academic community, such as Library Trends, Jasis/Jasist, Communication ACM, Information Processing & Management, Scientometrics, College Research Library, Journal of Documentation and Science.
77By investigating the index of cited documents, 55,753 documents titles were compiled (journal articles and books), totalizing 63,155 general citations from which 84 documents with over ten citations are considered on Table 6.
78The frequency of documents citations is consistent with the index of the most representative authors in their respective areas, according to the analysis of authors citations’ frequency.
Table 6: Frequency of documents citation (WoS 1994-2008)
Documents cited
|
Fq
|
Documents cited
|
Fq
|
Documents cited
|
Fq
|
Salton G, 1983, Intro Modern Informa
|
77
|
Barzilai G, 2003, Communities Law Poli
|
12
|
Marshall P, 1997, Res Methods How Desi
|
11
|
Saracevic T, 1988, J Am Soc Inform Sci, V39, P197
|
76
|
Blair Dc, 1990, Language Representat
|
12
|
Meley Ms, 1994, Datapro, V3
|
11
|
Salton G, 1989, Automatic Text Proce
|
44
|
Brin S, 1998, Comput Networks Isdn, V30, P107
|
12
|
Morgan G, 1983, Method Strategies So
|
11
|
Salton G, 1990, J Am Soc Inform Sci, V41, P288
|
35
|
2002, Fed Reg 0222, V67, P8452
|
11
|
Nicholas D, 2002, Nhs Direct Online In
|
11
|
Salton G, 1988, Information Processi, V24, P513
|
34
|
Bates Mj, 1979, J Am Soc Inform Sci, V30, P205
|
11
|
Orlikowski Wj, 2002, Organ Sci, V13, P249
|
11
|
Baezayates R, 1999, Modem Information Re
|
32
|
Bates Mj, 1990, Inform Process Manag, V26, P575
|
11
|
Peterson Ra, 1997, J Acad Market Sci, V25, P329
|
11
|
Belkin Nj, 1982, J Doc, V38, P61
|
30
|
Benjamin Ri, 1990, Long Range Plann, V23, P29
|
11
|
Quillin Tj, 2000, Ind Button
|
11
|
Rogers Em, 1986, Communication Techno
|
22
|
Blumenfeld L, 2003, Washington Post 0708, A1
|
11
|
Rivest R, 1987, Mach Learning, V2, P229
|
11
|
Belkin Nj, 1992, Commun Acm, V35, P29
|
20
|
Broadypreston J, 1998, Int J Inform Manage, V18, P277
|
11
|
Ryan S, 2000, Virtual U Internet R
|
11
|
Bilal D, 2001, J Am Soc Inf Sci Tec, V52, P118
|
20
|
Caldwell B, 1998, Mill Mess Inf W 0413
|
11
|
Scheiderman B, 1992, Designing User Inter
|
11
|
Borgman Cl, 1996, J Am Soc Inform Sci, V47, P568
|
20
|
Callan Jp, 1995, P 18 Ann Int Acm Sig, P21
|
11
|
Shneiderman B, 1998, Designing User Inter
|
11
|
Salton G, 1971, Smart Retrieval Syst
|
20
|
Chen Hc, 1995, J Am Soc Inform Sci, V46, P194
|
11
|
Spender Jc, 1996, Strategic Manage J, V17, P45
|
11
|
Schamber L, 1994, Annu Rev Inform Sci, V29, P33
|
19
|
Conklin J, 1987, Ieee Comput, V20, P17
|
11
|
Sullivan M, 2003, Keyword Magic Truth, P1
|
11
|
Taylor Rs, 1968, Coll Res Libr, V29, P178
|
17
|
Currie Dl, 1995, J Agr Food Informati, V3, P13
|
11
|
Thompson Jd, 1967, Org Action
|
11
|
Barry Cl, 1994, J Am Soc Inform Sci, V45, P149
|
16
|
Denvir Dp, 2003, Academe, V89, P36
|
11
|
Vandijk L, 2000, Digitalisering Delee
|
11
|
Bates Mj, 1989, Online Rev, V13, P407
|
16
|
Dumais St, 1997, Aaai 97 Spring S Cro, P18
|
11
|
Warner J, 2000, Aslib Proc, V52, P350
|
11
|
Berry Mw, 1999, Siam Rev, V41, P335
|
16
|
Euster Jr, 1984, P 3 Nat C Ass Coll R
|
11
|
Willcocks L, 1994, Information Manageme, P49
|
11
|
Buckland Mk, 1992, J Am Soc Inform Sci, V43, P284
|
16
|
Foucault M, 1970, Order Things Archaeo
|
11
|
Yang Y, 1999, J Information Retrie, V1, P9
|
11
|
Blair Dc, 1985, Commun Acm, V28, P289
|
15
|
Gastwirth Jl, 1972, Rev Econ Stat, V54, P306
|
11
|
Yang Y, 1999, P 22 Ann Int Acm Sig, P42
|
11
|
Braun T, 1994, Scientometrics, V31, P3
|
15
|
Graham G, 1999, Financial Times 0113
|
11
|
Zobel J, 1995, Inform Process Manag, V31, P361
|
11
|
Salton G, 1968, Automatic Information
|
14
|
Hammer M, 1990, Harvard Bus Rev, P103
|
11
|
Zuboff S, 1989, Age Smart Machine
|
11
|
Zipf Gk, 1949, Human Behav Principl
|
14
|
Heijden H, 2003, Information Manageme, V40, P541
|
11
|
Almind Tc, 1997, J Doc, V53, P404
|
10
|
Bates Mj, 1993, Libr Quart, V63, P1
|
13
|
Hoot Jl, 1983, Educ Gerontol, V9, P493
|
11
|
Belkin Nj, 1980, Canadian J Informati, V5, P133
|
10
|
Callan Jp, 1994, P 17 Ann Int Acm Sig, P302
|
13
|
Kaufman Mt, 2002, Soros Life Times Mes
|
11
|
Belkin Nj, 1990, J Inform Sci, V16, P11
|
10
|
Zipf Gk, 1949, Human Behav Principl
|
13
|
Koehler W, 2002, Ref User Serv Q, V42, P41
|
11
|
Belkin Nj, 1993, Inform Process Manag, V29, P325
|
10
|
Rogers Em, 1995, Diffusion Innovation
|
13
|
Lancaster Fw, 1993, If You Want Evaluate
|
11
|
Beniger Jr, 1986, Control Revolution T
|
10
|
1997, Times Higher Ed 0103, P12
|
12
|
Lewin K, 1951, Field Theory Social
|
11
|
Salton G, 1975, Commun Acm, V18, P613
|
10
|
Allen Tj, 1977, Managing Flow Techno
|
12
|
Lucier C, 1997, Strategy Business, P14
|
11
|
Salton G, 1983, Commun Acm, V26, P1022
|
10
|
NOTE : Fq = Frequency.
79On Figure 5, we observe documents published by the most relevant authors in each thematic areas, such as SALTON G, 1983, INTRO MODERN INFORMA (information retrieval), with a frequency of 77 appearances; SARACEVIC T, 1988, J AM SOC INFORM SCI, v39, p197, 76 appearances and BELKIN NJ, 1982, J DOC, v38, p 61, 30 appearances, being configured as fundamental authors for the consolidation of the Information Science area. The authorship of the most cited documents is consistent with the index of the most influent authors in Information Science.
Figure 5: Co-citation of documents (WoS 1994-2008)
80Figure 5 allows us to visualize the centrality of the most representative authors in their respective thematic area. It is observed that Salton appears as a central author and so are all other authors who are circled in pink, such as Lotka, White, Garfield and Egghe for metric studies, and Belkin for grounding of Information Science.
81Databases are facilitating the scientific knowledge access through indexation of scientific periodicals in their repertoire. It appears that some international titles represented in Web of Science, are among the most important for Information Science.
82Regarding the authors productivity, it is verified that the most productive authors are the most relevant in their respective thematic areas, showing relevant citation indexes, such as Salton, in information retrieval, and Belkin in grounding of Information Science.
83According to Bradford’s theory, some of the journals that present high citation indexes are: Journal of the American Society Information Science, Information processing & Management, College Research Library, Communication ACM and Journal of Documentation.
84According to Lotka’s law, Tenopir, Nicholas, Oppenheim, Oder, Rousseau and Egghe are among the highest producers of the field.
85Based on Zipf’s law, the keywords frequency has been identified and it has been observed that the compound terms are the most representative for Information Science area.
86It has been found that only two of the most renowned journals in Information Science area are mentioned in the Qualis/Capes Brazilian list. The results of our analysis suggest that the Qualis/Capes titles’ list might not be covering the best journals for the area.
87Building on our analysis of the information science area, we consider the following actions should be taken:
-
First, Universities should be aware of the importance of periodicals indexed in the Web of Science for the Information Science area, encouraging students and professors to promote those titles;
-
Second, there should be a rearrangement of the Qualis/Capes system, replacing the current criterion based on journals in which the professors publish each year, which is believed not to be a suitable quality criterion. Indeed, there are journals of low or fake quality that are on the list just because some professors publish in them ;
-
Third, professors and students should be encouraged to publish in more representative and renowned titles in Information Science.
88Publications in Brazil, in the area of Information Science, has a local feature, especially because of the Qualis / Capes system. According to us, this local feature turns the Brazilian publications difficult to be internationalized and studied.