iMechanica - Peer review - Comments

Lets also rank journals / I start with IJSS who wants to do more

Mike Ciavarella — Thu, 17 Jul 2008 08:52:08 -0400

I just did IJSS

Most cited papers and H-factor of some mechanics journals -- IJSS

who wants to do JMPS, and the other journals?

Mike

Is there a way to measure contribution of each ref. on a paper?

RoozbehSanaei — Wed, 16 Jul 2008 10:32:45 -0400

Is there any scientometric way to calculate a contribution of each ref. on a paper? it may be important.

more Citation metrics from Publish or Perish User's Manual

Mike Ciavarella — Wed, 16 Jul 2008 07:04:49 -0400

Citation metrics

Publish or Perish User's Manual

Citation metrics

Publish or Perish calculates the following citation metrics:

Total number of papers
Total number of citations
Average number of citations per paper
Average number of citations per author
Average number of papers per author
Hirsch's h-index and related parameters, shown as
h-index and Hirsch a=y.yy, m=z.zz in the output
Egghe's g-index, shown as g-index in the output
The contemporary h-index, shown as hc-index
and ac=y.yy in the output
Two variations of the individual h-index, shown as
hI-index and hI,norm in the output
The age-weighted citation rate
An analysis of the number of authors per paper.

Please note that these metrics are only as good as their input. We recommend
that you consult the following topics for information about the limitations of
the citation metrics and the underlying sources that Publish or Perish uses:

h-index

The h-index was proposed by J.E. Hirsch in his paper An index to quantify
an individual's scientific research output, arXiv:physics/0508025 v5 29 Sep 2005. It is defined as
follows:

A scientist has index h if h of his/her Np papers have at least h
citations each, and the other (Np-h) papers have no more than h
citations each.

It aims to measure the cumulative impact of a researcher's output by looking
at the amount of citation his/her work has received. Publish or Perish
calculates and displays the h index proper, its associated
proportionality constant a (from Nc,tot =
ah2), and the rate parameter m (from h ~ mn, where
n is the number of years since the first publication).

The properties of the h-index have been analyzed in various papers; see for
example Leo Egghe and Ronald Rousseau: An informetric model for the
Hirsch-index, Scientometrics, Vol. 69, No. 1 (2006), pp. 121-129.

This metric is shown as h-index and Hirsch a=y.yy,
m=z.zz in the output.

g-index

The g-index was proposed by Leo Egghe in his paper Theory and practice of
the g-index, Scientometrics, Vol. 69, No 1 (2006), pp. 131-152. It is
defined as follows:

[Given a set of articles] ranked in decreasing order of the number of
citations that they received, the g-index is the (unique) largest number such
that the top g articles received (together) at least g2
citations.

It aims to improve on the h-index by giving more weight to highly-cited
articles.

This metric is shown as g-index in the output.

Contemporary h-index

The Contemporary h-index was proposed by Antonis Sidiropoulos, Dimitrios
Katsaros, and Yannis Manolopoulos in their paper Generalized h-index for
disclosing latent facts in citation networks, arXiv:cs.DL/0607066
v1 13 Jul 2006.

It adds an age-related weighting to each cited article, giving (by default;
this depends on the parametrization) less weight to older articles. The
weighting is parametrized; the Publish or Perish implementation uses
gamma=4 and delta=1, like the authors did for their experiments.
This means that for an article published during the current year, its citations
account four times. For an article published 4 years ago, its citations account
only one time. For an article published 6 years ago, its citations account 4/6
times, and so on.

This metric is shown as hc-index and ac=y.yy in
the output.

Individual h-index (2 variations)

The Individual h-index was proposed by Pablo D. Batista, Monica G. Campiteli,
Osame Kinouchi, and Alexandre S. Martinez in their paper Is it possible to
compare researchers with different scientific interests?,
Scientometrics, Vol 68, No. 1 (2006), pp. 179-189.

It divides the standard h-index by the average number of authors in the
articles that contribute to the h-index, in order to reduce the effects of
co-authorship; the resulting index is called hI.

Publish or Perish also implements an alternative individual h-index,
hI,norm, that takes a different approach: instead of dividing the
total h-index, it first normalizes the number of citations for each paper by
dividing the number of citations by the number of authors for that paper, then
calculates hI,norm as the h-index of the normalized citation
counts. This approach is much more fine-grained than Batista et al.'s; we
believe that it more accurately accounts for any co-authorship effects that
might be present and that it is a better approximation of the per-author impact,
which is what the original h-index set out to provide.

These metrics are shown as hI-index (Batista et al.'s) and
hI,norm (PoP's) in the output.

Age-weighted citation rate (AWCR, AWCRpA) and AW-index

The age-weighted citation rate was inspired by Bihui Jin's note The
AR-index: complementing the h-index, ISSI Newsletter, 2007, 3(1), p.
6.

The AWCR measures the number of citations to an entire body of work, adjusted
for the age of each individual paper. It is an age-weighted citation rate, where
the number of citations to a given paper is divided by the age of that paper.
Jin defines the AR-index as the square root of the sum of all age-weighted
citation counts over all papers that contribute to the h-index. However, in the
Publish or Perish implementation we sum over all papers instead, because we feel
that this represents the impact of the total body of work more accurately. (In
particular, it allows younger and as yet less cited papers to contribute to the
AWCR, even though they may not yet contribute to the h-index.)

The AW-index is defined as the square root of the AWCR to allow comparison
with the h-index; it approximates the h-index if the (average) citation rate
remains more or less constant over the years.

The per-author age-weighted citation rate is similar to the plain AWCR, but
is normalized to the number of authors for each paper.

These metrics are shown as AWCR, AWCRpA and
AW-index in the output.

Publish or Perish User's Manual - Topic last modified on 27/01/08 12:19
Copyright © 1990-2008 Tarma Software
Research. All rights reserved.

Yes, I read it monthly and also Nature comments in this regard.

RoozbehSanaei — Mon, 14 Jul 2008 06:04:51 -0400

that is a nice journal, i think it discusses an important issue despite it's low IF!. I also read Nature comments in this regard. most of papers are based in this assumption : "Authors choose their references!" but in fact " authors just use any hing that they can use it faster and justifies their work better. i think this picture from PHD comics is near to reality and also funny. even best scientists do like this for those papers which are not very related to their main topic. It does not mean scientists are doing weak works, It means bibliometrics are not very trustful.

if you are interested in this topic IMU report on citation statistics is also very informative.

Roozeh, but have you seen Scientometrics?

Mike Ciavarella — Mon, 14 Jul 2008 02:17:59 -0400

Scientometric is a journal full of this type of discussions. You should check it.

Only looking at one issue, e.g.

Scientometrics

Volume: 76, Issue: 1

July 2008

Save Selected |
Clear All

The evaluation of university departments and their scientists: Some
general considerations with reference to exemplary bibliometric
publication and citation analyses for a Department of psychology

Krampen, Günter
pp. 3 - 21

Bibliographic Page
| Article Full Text PDF (389 KB)

Scientometric definition of science: In what respect is the humanities
more scientific than mathematical and social sciences?

Eto, Hajime
pp. 23 - 42

Bibliographic Page
| Article Full Text PDF (240 KB)

Relation of seminal nanotechnology document production to total nanotechnology document production — South Korea

Kostoff, Ronald N.; Barth, Ryan B.; Lau, Clifford G. Y.
pp. 43 - 67

Bibliographic Page
| Article Full Text PDF (970 KB)

Great expectations: The role of Open Access in improving countries’ recognition

Sotudeh, Hajar; Horri, Abbas
pp. 69 - 93

Bibliographic Page
| Article Full Text PDF (304 KB)

Ranking Taiwanese management journals: A case study

Kao, Chiang; Lin, Hsiou-Wei; Chung, San-Lin; Tsai, Wei-Chi; Chiou, Jyh-Shen; Chen, Yen-Liang; et. al.
pp. 95 - 115

Bibliographic Page
| Article Full Text PDF (258 KB)

The mathematical relation between the impact factor and the uncitedness factor

Egghe, Leo
pp. 117 - 123

Bibliographic Page
| Article Full Text PDF (153 KB)

How do we measure the use of scientific journals? A note on research methodologies

Moghaddam, Golnessa Galyani; Moballeghi, Mostafa
pp. 125 - 133

Bibliographic Page
| Article Full Text PDF (156 KB)

PageRank for bibliographic networks

Fiala, Dalibor; Rousselot, François; Ježek, Karel
pp. 135 - 158

Bibliographic Page
| Article Full Text PDF (601 KB)

The delineation of nanoscience and nanotechnology in terms of journals and patents: A most recent update

Leydesdorff, Loet
pp. 159 - 167

Bibliographic Page
| Article Full Text PDF (434 KB)

What makes an article influential? Predicting impact in social and personality psychology

Haslam, Nick; Ban, Lauren; Kaufmann, Leah; Loughnan, Stephen; Peters, Kim; Whelan, Jennifer; et. al.
pp. 169 - 185

Bibliographic Page
| Article Full Text PDF (217 KB)

The influence of self-citation corrections on Egghe’s g index

Schreiber, Michael
pp. 187 - 200

Bibliographic Page
| Article Full Text PDF (309 KB)

Thanks for your link.

RoozbehSanaei — Mon, 14 Jul 2008 01:13:10 -0400

This link is very useful at least for me. Time shortage forces people to use numbers.research evaluation enhances like any other methodologies. Dynamic review paper also is a way to judge about different works in a cooperative manner.

Re: Google scholar and h-index (aka Google APIs)

Biswajit Banerjee — Mon, 14 Jul 2008 00:46:23 -0400

Google has provided several APIs (Application Programming Interfaces) for outsiders to use Google's data.

There's a Google project called citations-gadget which may be what you're looking for. You can find other APIs at http://code.google.com/

My own views on h-index have been summed up in an excellent blog post by Bee (of BackReaction) called

We have only ourselves to judge on each other.

To quote Bee

"Don't fall for shortcuts. Under time pressure, people tend to look for easy criteria to judge on other people or their projects. Such might be the citation index, the number of papers, prominent coauthors or grants obtained. In all instances scientists should judge on other's work by themselves. This requires that scientists have enough time for this judgement. Time is essential, but too short in today's research atmosphere. For a thought on how to cut back on researcher's duties, see previous post on Research and Teaching."

-- Biswajit

F. Moisy on how we can generalize multiple authors..

Mike Ciavarella — Sun, 13 Jul 2008 16:13:26 -0400

Dear M. Ciavarella

This should work for a list of people:

name={'a einstein','f moisy','c bruni'};
for n=1:length(name),
h(n)=hindex(name{n});
disp([name{n} ' ' num2str(h(n))]);
end

Sincerely
F. Moisy

For example, by cleaning "M.Ciavarella" I find H=13

Mike Ciavarella — Sun, 13 Jul 2008 15:49:07 -0400

The number of articles is 0

The total number of citations is 624

The maximum number of citations is 56

The h-number (or h-index) is 13

The m-number (or m-index) is 0.34210526315789475 (1970-2008)

but then the output is very slow....both in web1 or web2 format. not sure why!

But I start to think the best is this french one...

Mike Ciavarella — Sun, 13 Jul 2008 15:35:16 -0400

I start to think the best sofware is this from collegues in Paris --- maybe I can contact them!

http://hview.limsi.fr/

In fact, I suspect you can use authomatic scripts by using for example the string

http://hview.limsi.fr/hpagew2.php?level=1&subj=chm&subj=eng&name=%22M.Ci...

or maybe multiple authors so we can play with. The problem with Google Scholar is that it seems it is NOT classified by LOCATION or INSTITUTION, but only AUTHOR and FIELD.

However, see the help:

Help

Use the OR operator with quotes for names with non-unique spellings; for example,
The guy you think to know as Fred Vernier should be writen:

"Frédéric Vernier" OR "Frederic Vernier"
On the other hand you still miss some of his articles. If you type
"F Vernier" you will have all his articles but ... also articles from someone else. You then have to deselect
them is order to find the right h-curve !
Warning Restricting an author to a given domain will remove all the articles
google scholar couldn't classify ! For instance articles not written in english. Restricting to given years
will also remove all articles google scholar couldn't determine the year.
Most of the time data must be cleaned ! remove articles from others and
merge articles (when typos made google duplicate a single article)
We disclaim everything, especially every problem resulting from the data returned by Google Scholar.
Data you enter is not logged yet ... but we want to save all
the cleaning work done by users in order to improve the data source !
You may save the url in the address textfield of your browser but we
will change the format soon and the old address will return an error.
You may enter many names in the author textfield to evaluate a full team of researsher but the system is already slow ...
You may experience denial of service from google scholar
sometime. Retry 2 or 3 times. If you still have nothing come back the
day after.
I did compute the h-view of the whole UIST conference and the first 3 for the 20th anniversary of UIST. The poster was presented at UIST'07.

No, at the moment I am trying to understand QuadSearch

Mike Ciavarella — Sun, 13 Jul 2008 15:26:52 -0400

About QuadSearch

The metasearch
engines are web services designed to transfer the user's queries to
multiple existing search engines. A metasearch engine does not maintain
its own index of documents. It collects and reorganizes the result
lists (top-k lists), then it returns the processed data to the user.

Compared to a classic single search engine, a metasearch engine offers:

Increased web coverage
Improved retrieval effectiveness
Effortless invocation of multiple search engines

QuadSearch is a brand new and constantly evolving metasearch engine. It is hosted by the Raptor web server, which belongs to the Data Engineering Laboratory (DELAB), of Aristotle University of Thessaloniki.

It has been created by Leonidas Akritidis, George Voutsakelis, Dimitrios Katsaros and Panayiotis Bozanis.

Their
main purpose is to build a powerful scientific tool, that will be used
by the academic community in order to retrieve useful information.
Their secondary objective focuses on finding better and faster ranking
algorithms.

QuadSearch's heart is
the KE Algorithm, which is the mechanism that determines the
classification of the collected results. KE Algorithm was invented by
the founders of this engine and has also been the basis for Search Engine Analyzer.
Apart from KE Algorithm, QuadSearch is capable of using a variety of
several other ranking algorithms, to classify the results.

As mentioned
above, QuadSearch is constantly under construction. If you have any
suggestions for it, or you want to report a security issue, please let
us know, by filling the appropriate contact form.

I think h factor for subjects is even more rational.

RoozbehSanaei — Sun, 13 Jul 2008 13:31:04 -0400

I think h factor for subjects is even more rational than h factor for authors.

personally i prefer to compare citation number of a paper with citation number of refered papers for evaluating a paper in a subject. i think it is very rational.

I am trying with JScript now.

RoozbehSanaei — Sun, 13 Jul 2008 13:28:04 -0400

I am trying with Java Script.

[I like JScript because it's free, works in all versions of Windows, linux, Mc and even handphone!]

Please make me aware if you got more result with matlab.

Ok, one or the other, let's do it -- we cannot only talk about..

Mike Ciavarella — Sun, 13 Jul 2008 11:43:08 -0400

michele ciavarella
www.micheleciavarella.it

However, this code does NOT seem to work !

Mike Ciavarella — Sun, 13 Jul 2008 05:22:39 -0400

I get an error which seems from the matlab web site that had appeared previously, but maybe it is worse? google scholar maybe interprets any call from matlab as an attack even for a single use!

If I go to
http://scholar.google.com/scholar?num=100&hl=it&lr=&q=Albert+Einstein

it works

but if I write
urlread('http://scholar.google.com/scholar?num=100&hl=it&lr=&q=Albert+Einstein')

there is the error

whereas if I write plain

urlread('http://scholar.google.com/')

it works again.

How to solve this?

Regards, Mike