This shows you the differences between two versions of the page.

Link to this comparison view

vlado:pub:sreda:1214 [2012/06/18 12:03] (current)
vlado created
Line 1: Line 1:
 +====== 1214. sredin seminar, 20. junij 2012 ======
 +Gomez Nuñez, Antonio Jesus <antoniojesus.gomez@cchs.csic.es>
 +**Improving the Categorization of Scopus Journals included in SCImago Journal & Country Rank (SJR)**
 +Scientific information stored in large scientific multidisciplinary
 +databases requires a good organization and arrangement not only for
 +information retrieval purposes but for developing reliable and
 +non-misleading indicators about impact, collaboration, visibility...,
 +within disciplines like Bibliometrics and Scientometrics. Similarly, a
 +good classification of information, regardless of aggregation level,
 +is desirable for information visualization or network analysis, whose
 +main surveys are based on information covered by scientific databases.
 +Among the most prestigious and remarkable ones are Web of Knowledge
 +(WOK) [Thomson Reuters] and Scopus [Elsevier]. Both use a similar
 +classification scheme according to a hierarchical system in two levels
 +composed of areas (broad level) and categories (specific level).
 +Among different tools for analysis and assessment of scientific
 +information "SCImago Journal & Country Rank is a portal that includes
 +the journals and country scientific indicators developed from the
 +information contained in the Scopus® database (Elsevier B.V.). These
 +indicators can be used to assess and analyze scientific domains"
 +(Scimago Lab. http://www.scimagojr.com/). Starting from the previous
 +classification of journals produced by Scopus, the categorization of
 +journals was refined following different criteria like opinion of
 +experts, tiles and scopes of journals.
 +Hereupon, it was pretended to improve and to tune the categorization
 +of the SJR journal set using automatic and statistical procedures, or
 +at least, avoiding the human mediation so far as possible. Thus, a
 +first work to improve the classification scheme of SJR working from
 +initial categorization and using reference analysis in combination
 +with different citation thresholds to determine the final category of
 +every journal was implemented. This method showed a solid performance
 +in grouping journals at a level higher than categories —that is,
 +aggregating journals into subject areas. It also enabled us to
 +redesign the SJR classification scheme, providing for a more cohesive
 +one that covers a good proportion of re-categorized journals. Anyhow,
 +in order to obtain a better categorization of journals, the method
 +should be complemented with additional techniques.
 +For following work, it was decided to make clustering of journals
 +using a combination of three citation measures, namely, Direct
 +Citation (DC), Cocitation (CC) and Bibliographic Coupling (BC).
 +Using R statistical software, an asymmetrical journal-journal matrix
 +with the sum of fractionalized 3-citation-measures was constructed
 +and then, values were transformed into cosine similarities. In
 +closing, similarities values were transformed into distances and
 +Ward hierarchical clustering was applied on them.
 +The proposal to develop in Ljubljana is related to the use of
 +software Pajek and, concretely, island analysis to detect different
 +sub-networks (clusters) from the global journal citation network
 +formed by around 19000 Scopus journals.
 +For the future research, it seems interesting to employ new
 +statistical/automatic techniques or network analysis adopting a
 +combination of different variables like citation measures, text of
 +documents (title, abstract and/or keywords), or address of authors,
vlado/pub/sreda/1214.txt · Last modified: 2012/06/18 12:03 by vlado
Except where otherwise noted, content on this wiki is licensed under the following license: CC Attribution-Noncommercial-Share Alike 3.0 Unported
Recent changes RSS feed Donate Powered by PHP Valid XHTML 1.0 Valid CSS Driven by DokuWiki