2. Abstracting & Indexing| 10/7/2014 | 2
Overview
• Thomson Reuters (ISI)
• PubMed/Medline/PubMed Central
• Google Citation Database
• Microsoft Academic Search
• SCOPUS/EI Compendex
• Other A&I Services
• Other Ranking Systems
• Evaluating and Improving Journal Quality
3. Abstracting & Indexing| 10/7/2014 | 3
Thomson Reuters (ISI)
• ISI Application and Evaluation
• ISI Databases
• ISI Indexing
• ISI Proceedings Index
• ISI Book Index
• Impact Factor Definition
• Improving the Impact Factor
4. Abstracting & Indexing| 10/7/2014 | 4
Application
• Prerequisites
• Titles, keywords, and abstracts in English, author names in roman script
• Reference lists in Roman script
• Application submission
• For journals published by Springer, submission through Dr. Guenther Eichhorn
• For journals published elsewhere, submission directly to ISI
• Description of the journal in the application
• Show why the journal is unique, what differentiates it from other journals
• Show the importance of the Editorial Board
• Describe the editorial process and how you handle ethical problems, etc.
• Show how international the journal is (authors, readers, EB)
• Letters of recommendation
• Describe why the journal is important for the scientific community
• Recommend indexing by ISI, since the journal is important to the community
• Do NOT say that ISI ‘should’ include the journal, do NOT tell ISI what they should do
• Do NOT say that the journal will benefit from being included
5. Abstracting & Indexing| 10/7/2014 | 5
During Evaluation
• Evaluation of journals published by Springer
• Evaluation of electronic version from SpringerLink
• Evaluation of journals published elsewhere
• Evaluation of electronic version if available, otherwise provide print subscription
• Timeliness
• Issues need to appear online on time, according to the cover date of the issues
• For consecutive article publishing, they need to be published regularly without
larger gaps in time (at least about two per month)
• Number of articles
• ISI expects at least 24 articles per year.
• Criteria
• Number of citations to the journal
• Ranking of the journal in its category
• Number of citations to authors publishing in the journal
• Number of citations to members of the editorial board
• Internationality of authors
• Internationality of members of the editorial board
6. Abstracting & Indexing| 10/7/2014 | 6
Results of Evaluations
• Accepted
• ISI determines into which database(s) to include the journal.
• ISI usually (but not always) includes the last couple of years so the journal can get
an Impact Factor for the year of acceptance
• Journals that are accepted within the first 3 volumes, are included from volume 1
• Reasons for Rejection
• Not published on time
• Low citation rate, low ranking in category
• Too few articles (at least 24 per year are required)
• Not enough scientific quality
• Authors and/or board members not cited enough
• Too many proceedings
• Subject area already covered by other ISI journals
• Resubmit Schedule after Rejection
• 3 years after 1st rejection, 6 years after 2nd rejection, 9 years after 3rd rejection
7. Abstracting & Indexing| 10/7/2014 | 7
ISI Databases (Web of Science)
• Science Citation Index Expanded (SCI-E)
• Science Citation Index (SCI). Subset of SCI-E
• Must be in top 10% of SCI-E categories to apply
• Social Science Citation Index (SSCI)
• Arts & Humanities Citation Index (A&HCI), no Impact Factor
• Conference Proceedings Citation Index- Science (CPCI-S), no Impact Factor
• Conference Proceedings Citation Index- Social Science & Humanities (CPCI-SSH)
• Book Citation Index, Science (BKCI-S), no Impact Factor
• Book Citation Index, Social Sciences & Humanities (BKCI-SSH), no Impact Factor
• Journal Citation Report (JCR). Publishes the Impact Factors
• Other discipline-specific databases (e.g. BIOSIS, Zoological Abstracts, etc)
8. Abstracting & Indexing| 10/7/2014 | 8
ISI Indexing
• Indexing at ISI has recently experienced workflow challenges and delays
• Monitor your journal in ISI
• All indexing now done from electronic data, if available
• Currently indexing from PDF, not electronic XML data
• They expect to start using full text XML data this year
• Traditional, issue based journals
• Indexed when issue appears online
• Online-First articles are not included, since they do not have final page numbers,
but their citations count for the Impact Factor, as long as the citation contains the
journal name and the year of publication
• Consecutive article publishing (CAP)
• Articles are included as they are published. Articles must have final form at the
time of publication.
• Do not assign page numbers later (e.g. for a summary volume)
9. Abstracting & Indexing| 10/7/2014 | 9
ISI Proceedings Database
• All proceedings books published by Springer go to ISI automatically for both
individual proceedings and proceedings series
• Individual decisions for each volume, no collective advanced decisions for
proceedings series
• Proceedings issues from journals are duplicated into the proceedings database
• No Impact Factor for Proceedings database.
• ISI frequently misses proceedings volumes, they sometimes need to be reminded
about proceedings volumes.
• New procedures for alerting ISI to new proceedings volumes published by Springer
are in place. This will hopefully improve the coverage of proceedings by ISI.
• ISI limits the number of proceedings volumes that they index to 4,000 per year for all
publishers combined.
10. Abstracting & Indexing| 10/7/2014 | 10
ISI Book Index
• Released on 9 October 2011
• All new Springer books are sent to ISI automatically
• Evaluated mostly from electronic version. Large, multi-volume works are evaluated
from print
• Evaluation
• Each book is evaluated individually, even if it is published as part of a series.
• Selection Criteria
• Books with copyright of 2005 or later are considered
• Consist of scholarly articles with complete reference lists
• Limit of 10,000 books per year for all publishers combined
• List of accepted books online
• Currently 90% acceptance rate for eligible books
• Over 8000 Springer books accepted
• Check for accepted books: http://wokinfo.com/mbl/
• Book Index Properties
• No Impact Factor or other ranking. Books don’t fit into the Impact Factor algorithm.
• Citations from books to journals do NOT count for the journal’s Impact Factor
11. Abstracting & Indexing| 10/7/2014 | 11
ISI Book Index (cont’d)
• Reasons for rejection
• Popular titles
• Textbooks/”Introductory” texts (post-graduate text books are eligible)
• Copyright date out of scope (only copyright of 2005 or later considered)
• Reprint of previous editions/reprints of journal articles
• Manual/handbook/guide/dictionary/atlas
• Dissertations/theses
• Proceedings (these are all considered for the Proceedings database)
• No citations or references/not scholarly
• ISI frequently misses books, they sometimes need to be reminded about books.
• New procedures for alerting ISI to new Springer books are in place. This will
hopefully improve the coverage of books by ISI.
12. Abstracting & Indexing| 10/7/2014 | 12
Legacy Book Series in SCI-E and SSCI
• ISI did include some book series in their citation database. These are all series that
were in SCI-E or SSCI before they started their book index. They do no longer accept
book series in SCI-E or SSCI, regardless of how well cited they are.
13. Abstracting & Indexing| 10/7/2014 | 13
Impact Factor Definition
• Average number of citations from articles published in the Impact Factor year to
articles published in the previous two years. E.g.:
• 20 articles published in 2011
• 30 articles published in 2012
• 100 citations from articles published in 2013 in any journal in WoS to articles
published in 2011 and 2012 in your journal
• Impact Factor(2013) = 100/(20+30) = 2.000
14. Abstracting & Indexing| 10/7/2014 | 14
Improving the Impact Factor
• Publish some high quality review articles
• Publish methodology articles
• Publication timing
• Important articles (like review and methodology articles) and special issues should
be published at the beginning of the year, so they have more time to accumulate
citations.
• Solicit articles by highly cited authors and on highly cited subjects
• Analyses can be done through Web of Science. Templates for such analyses and a
document that describes techniques to evaluate and improve a journal are
available.
• Special issues on hot topics can help
• Promote special issues at conferences by including reprints or whole issues in
conference packages
• Encourage citations to other relevant articles in your journal
• Be careful, too many self citations will result in removal from the JCR.
• Promote your journal at conferences
• Some conferences allow you to insert reprints in their conference handouts.
15. Abstracting & Indexing| 10/7/2014 | 15
Dangers
• Too many self citations. If a journal has too many self citations, ISI will remove the
journal from the JCR for initially two years. If the situation does not improve, ISI may
permanently remove the journal from the JCR.
• ISI does not give firm threshold values, here are guidelines.
• >20% self citations is the danger zone.
• >50% almost guarantees removal
• Too many proceedings. ISI removes journals from the SCI-E or SSCI if they publish
too many proceedings. One proceedings volume per year should be the maximum.
If the journal has 50% proceedings articles, it is in grave danger of being removed
from the SCI-E.
16. Abstracting & Indexing| 10/7/2014 | 16
PubMed/Medline/PubMed Central
• PubMed: Freely searchable database at the National Library of Medicine
• Medline: Subset of data in PubMed of journals that have been reviewed and
accepted for full indexing
• PubMed Central: Collection of Open Access and Free Access articles. Metadata for
these articles go into PubMed (but not Medline)
• Difference between articles in Medline and articles only in PubMed:
• Medline articles are fully indexed with MeSH indexing keywords. Errata are
processed and applied as appropriate, funding numbers are extracted and indexed.
17. Abstracting & Indexing| 10/7/2014 | 17
Medline
• New policy for e-only journals accepted in Medline
• Need to be deposited in permanent archives
• Springer deposits all journals for which we have permission in Portico and CLOCKSS
• Need to be available online at NLM
• All Springer e-only journals are available to NLM with their subscriptions
• Need to be delivered to PubMed as PDF/A
• PubMed (but not Medline) accepts metadata for back issues of Medline journals
• Springer is in the process of depositing such back issues for Springer journals
18. Abstracting & Indexing| 10/7/2014 | 18
PubMed Central
• Prerequisites for getting into PMC
• Full Open Access or Free Access with embargo of no more than 12 months
• At least 15 articles published
• Full text XML in JATS format
• Policies for PMC Journals
• Metadata for Open Access journals without embargo can go directly into PubMed
(including online-first articles)
• Free Access journals with embargo are deposited by PMC into PubMed only after
final issues are out of embargo (no online-first articles included)
• PMC accepts only articles with full text XML
• Exceptions are back issues, for which XML metadata and PDF files are sufficient.
19. Abstracting & Indexing| 10/7/2014 | 19
PubMed/Medline Application
• Prerequisite
• Four review issues available at the time of application (double issues count as one).
• e-only journals must have published at least 40 articles and must have published for at
least one year
• Application for Springer journals through Springer A&I
• Peer review process description is important, journal description is very important
• Application for journals from other publishers directly to PubMed
• Conflict of Interest statement mandatory in each article
• Individual statements for each author with author names
• Within the regular article pages, just before the reference list
• Statements of Human and Animal Rights required for articles that report studies with
human or animal participants
• Statements of Informed Consent required for articles that report studies with human
participants
• 2-3 Letters of recommendation from prominent scientists not on the editorial board may
help (not from NIH scientists)
• 3 reviews/year (February, June, October). Application deadlines: 1 February for June
review, 1 June for October review, 1 October for February review
• 10 Springer journals per review (so send me the application early)
20. Abstracting & Indexing| 10/7/2014 | 20
PubMed Review Results
• Results available 6-8 weeks after the review, review reports 2-3 months after review
• Decisions are final, no reversal possible
• Reasons for Rejection
• Scientific quality of the articles
• Quality of Author/Institutions (# of publications, # of citations)
• Quality of Editorial Board (# of publications, # of citations)
• Editorial Board Independence (ties to industry)
• Lack of relevance to scientists, clinicians, educators
• Lack of strict external review process
• Rejection rate too low, too many commissioned articles
• Conflict of Interest/Ethics compliance statements lacking
21. Abstracting & Indexing| 10/7/2014 | 21
PubMed
• Sponsored supplements will not be index without Conflict of Interest statements in
each article, including editorials
• Individual statements for each author with author names
• Within the regular article pages, just before the reference list
• The supplement will appear briefly in PubMed and will then be hidden. Once
PubMed has verified that the supplement complies with the requirement, the
articles will be included again
• Corrections in PubMed
• If Springer PDF not affected by error: Post-publication correction and resend of
data, as long as the article has not been processed by PubMed
• Errata about metadata are applied directly to the data (e.g. name corrections)
• Errata about main part of the article are linked to the original article
22. Abstracting & Indexing| 10/7/2014 | 22
Google
• Most traffic to Springer.com and SpringerLink comes from Google
• Google Scholar indexes everything we have online
• Journals
• Proceedings
• Books (usually on chapter level)
• Publish or Perish useful for data analysis
• Google Scholar Author citation database
• Authors should set up their profile, claim their articles
• Citation information for authors
• Calculates h-index for authors
• Google Scholar Journal citation database with ranking
• Only need 100 articles in previous five years in order to be included
• h5-index for journals
• Better ranking than ISI. ISI ranks review journal much higher than research
journals
23. Abstracting & Indexing| 10/7/2014 | 23
h-index example diagram
The chart shows the article
distribution for a hypothetical
journal with five articles that have
17, 9, 6, 3, and 2 citations
respectively. The h-index is defines
as follows:
The h-index of a publication is the
largest number h such that at least
h articles in that publication were
cited at least h times each.
The diagram shows the h-index,
the h-core, and the h-median as
defined by Google. The h5-index
that Google uses for their ranking is
an h-index calculated for articles
from the last five years.
17
Number of Citations
9
6
3
2
18
16
14
12
10
8
6
4
2
0
1 2 3 4 5
h-core
h-index: 3
h-median: 9
24. Abstracting & Indexing| 10/7/2014 | 24
Microsoft Academic Search
• Search system
• Author citation database
• Authors can set up profile, claim their articles
• h-index for authors
• Citation history
• More features than Google author citation database
• Graphical analysis tools
25. Abstracting & Indexing| 10/7/2014 | 25
Scopus/EI Compendex
• Elsevier products
• More selective now
• Application form online
•Evaluation takes ½ - 1 year
•Need to publish for one year before applying
• Scopus
•Search system
•Citation database
•Journal ranking
•h-index
• SCImago ranking system
• Uses Scopus data
26. Abstracting & Indexing| 10/7/2014 | 26
Miscellaneous information
• Full Text Indexing
• Google (all journals and all books)
• Summon (all journals)
• ADS (Physics, Astronomy, Geosciences)
• INSPIRE (High Energy Physics)
• Electronic Data
• All A&I Services now accept electronic data
• Mostly discipline specific
• Journal Directories (e.g. Cabell’s, Ulrich’s, etc)
• Not A&I services. Information per journal, not per article.
28. Abstracting & Indexing| 10/7/2014 | 28
Altmetrics, Review after publication through Crowdsourcing
• Use social media content (Twitter, Facebook, blogs) to judge individual articles
• Use article download statistics
• Ranking, post publication review
• Proponents for altmetrics claim that the Impact Factor is easy to manipulate. In my
opinion, altmetrics are even easier to manipulate.
• Altmetrics only count mentions in social media, they don’t distinguish between good
and bad mentions. For instance an article that reports wrong results, which is
mentioned in many blogs, will have a high altmetrics rating. Similarly, an article that
reports about some dubious weight loss scheme, will have tremendous altmetrics
ratings.
• Not everything in this world should be put to a vote — even if we have the
technology to do so – expert peer review has its place and should not be thrown out
for crowd sourcing.
• The Limits of Crowdsourcing in the Scientific Disciplines
• Criticism of Altmetrics: http://www.dcscience.net/?p=6369