blog dds


Impact Factor of Computer Science Journals 2011

The Thomson Reuters Web of Knowledge has published the 2011 Journal Citation Reports. Following similar studies I performed in 2007, '08, '09, '10, and '11, here is my analysis of the current status and trends for the impact factor of computer science journals.

Let me start with an overview of the method I followed. I performed my analysis on the 465 journals appearing under the Web of Knowledge's computer science subject categories: artificial intelligence; cybernetics; hardware & architecture; information systems; interdisciplinary applications; software engineering; theory & methods. Note that journals may appear in many categories. In particular, there are many overlaps between the above categories and "electrical and electronic engineering" and "operations research and management science".

Don't take this report too seriously. The impact factor is notoriously prone to misuse; let's not make the situation worse. Also note that in computer science there are tens of prestigious conferences whose impact factor has been calculated to be higher than that of most journals. For a ranking of these conferences you may wish to consult the Conference Ranking Exercise performed by the Australian Government's, Excellence in Research for Australia body.

Journals with Highest IF

2IEEE T PATTERN ANAL0162-88284.908
3J CHEM INF MODEL1549-95964.675
4SIAM J IMAGING SCI1936-49544.656
5ACM COMPUT SURV0360-03004.529
6MIS QUART0276-77834.447
7MED IMAGE ANAL1361-84154.424
8INT J NEURAL SYST0129-06574.284
9IEEE T FUZZY SYST1063-67064.260
10J STAT SOFTW1548-76604.010

This table demonstrates a tendency of impact factors to revert towards the mean, which for top computer science journals seems to be around 4.5. Thus, last year's 7.8 stratospheric impact factor of the ACM Computing Surveys has reverted to a more earthly 4.5. The same is the case for MIS Quarterly, which was slightly above 5 in 2010, and has returned back to 4.5 in 2011. I'm therefore betting that IEEE Communications Surveys and Tutorials won't keep its 6.3 impact factor next year.

Interestingly, two journals in the top-ten list are published only online: the IEEE Communications Surveys and Tutorials and the Journal of Chemical Information and Modeling. To me this shows that the taboo of paperless publishing has been broken.

Many of the entries in the top-ten list are probably benefiting from the higher impact factors associated with life science publications. This probably includes the five journals associated with pattern analysis methods and machine learning techniques, namely IEEE Transactions on Pattern Analysis and Machine Intelligence, SIAM Journal on Imaging Sciences, Medical Image Analysis, International Journal of Neural Systems, IEEE Transactions on Fuzzy Systems, the Journal of Chemical Information and Modeling and could well also include the Journal of Statistical Software.

Journals Added to the JCR in 2011

ACM T KNOWL DISCOV D1556-46811.419
ACM T RECONFIG TECHN1936-74060.650
APPL ONTOL1570-58381.105
COGN COMPUT1866-99561.000
FRONT COMPUT SCI CHI1673-73500.266
IEEE COMPUT ARCHIT L1556-60560.828
IEEE LAT AM T1548-09920.346
IEEE T AUTON MENT DE1943-06042.310
IEEE T COMP INTEL AI1943-068X1.617
IEEE T HAPTICS1939-14121.490
IEEE T LEARN TECHNOL1939-13820.823
IEEE T SERV COMPUT1939-13741.468
INT J DATA WAREHOUS1548-39240.200
INT J SENS NETW1748-12791.386
J APPL LOGIC1570-86830.574
J INTERNET TECHNOL1607-92640.508
J VISUAL-JAPAN1343-88750.435
NAT LANG ENG1351-32490.432

The 21 journals added to the list are a lower number than that of 2010 (24) and 2009 (a whooping 38) indicating the Thomson is adding journals at a conservative rate. This year IEEE made a big splash with eight new journals, demonstrating the power of the "build it and they will come" model. Also, as in last year there are journals serving a local audience: Revista IEEE América Latina and Frontiers of computer science in China. For the first time in five years no journal associated with IT security has been added to the list.

Journals Dropped from the JCR in 2011

NameISSNLast Year's IF
CONTROL CYBERN0324-85690.300
IBM SYST J0018-86701.792
INT J COMPUT INT SYS1875-68831.471
INT J INF TECH DECIS0219-62203.139
INT J INNOV COMPUT I1349-41981.664
INT J WAVELETS MULTI0219-69131.306

I was saddened to see the IBM System Journal depart from ISI's list, but it turns out that after 2008 the journal was merged with the IBM Journal of Research and Development. More surprising is the departure from the list of the International Journal of Information Technology and Decision Making with an impact factor of 3.1. One can only speculate about the reasons for this removal. It is well known that ISI is increasingly investigating and taking a hard line against publications that try to manipulate their impact factor, e.g. by encouraging authors to cite articles from previous issues.

Journals with the Highest Increase in IF

ENTERP INF SYST-UK1751-75752.898
IEEE T FUZZY SYST1063-67061.577
IEEE T IND INFORM1551-32031.363
J STAT SOFTW1548-76601.363
FOUND COMPUT MATH1615-33751.262
MOB INF SYST1574-017X1.107
COMPUT PHYS COMMUN0010-46550.968
ANNU REV INFORM SCI0066-42000.955
INT J WEB GRID SERV1741-11060.941

Again this year the list includes application-oriented journals, ranging from hard-core metal-bashing (IEEE Transactions on Industrial Informatics and Integrated Computer-Aided Engineering) to large scale systems (International Journal of Web and Grid Services and Enterprise Information Systems), to vertical application domains (Mobile Information Systems, Journal of Statistical Software, Computer Physics Communications). Computer science is increasingly becoming more practical and interdisciplinary.

Journals with the Highest Decrease in IF

IBM J RES DEV0018-8646-4.14
ACM COMPUT SURV0360-0300-3.277
HUM-COMPUT INTERACT0737-0024-2.524
COMPUT LINGUIST0891-2017-2.25
USER MODEL USER-ADAP0924-1868-1.674
EVOL COMPUT1063-6560-1.569
J AUTOM REASONING0168-7433-1.546
J WEB SEMANT1570-8268-1.487
J STRATEGIC INF SYST0963-8687-1.443
J CRYPTOL0933-2790-1.425

The two journals at the top of the list had very high impact factors, so this is probably a reversion towards the mean. More worrying is the appearance on the list of Human-Computer Interaction for a second year in a row. One can argue that a large correction is spanning two years. However, the co-existence of the journal User Modeling and User-Adapted Interaction in the list may also indicate the bursting of a bubble in HCI research, or that HCI researchers have given up scientific publishing to profit in the lucrative apps and games market.

Read and post comments, or share through   

Creative Commons License Last modified: Tuesday, July 3, 2012 3:43 pm
Unless otherwise expressly stated, all original material on this page created by Diomidis Spinellis is licensed under a Creative Commons Attribution-Share Alike 3.0 Greece License.