Template talk:Journal Info Template
From Eureka
| Antarctica Journal of Mathematics | |||||||||||||||
 | |||||||||||||||
| Basic Info | |||||||||||||||
| |||||||||||||||
| Editors | |||||||||||||||
| |||||||||||||||
| Impact (eigenfactor.org) | |||||||||||||||
| |||||||||||||||
| Price info (journalprices.com) | |||||||||||||||
| |||||||||||||||
| AMS (link) | |||||||||||||||
| |||||||||||||||
| |||||||||||||||
The Journal Info Template produces a box on the right hand side of a journal's page, showing the vital statistics of the journal. It provides links to other internal pages showing more information, as well as linking to external data sources such as www.eigenfactor.org.
When we upgrade our current installation of Mediawiki it will become much more powerful, due to the thereby-gained ability to obtain greatly improved functionality from the Winter extension and the latest version of Semantic Mediawiki.
Philosophy behind the template
The template is not the most important thing on a journal's page. Eureka is not just an aggregator of information on journals from various sources on the internet. I'd like to say that the most important thing on a journal's page is the human-written explanatory text. But that doesn't really hold water, since basic information on journals can indeed be found on Wikipedia.
I think that the template is simply something that makes Eureka a useful place for mathematicians to visit. Currently, our blurb on the front page says "We provide easily accessible information on scholarly publishing venues so that the competition among groups and strategies may be a fair one." I think this is fair, but personally, I think Eureka really takes a stand - a stand against corporate run monoopolies. Why should mathematicians produce material - not just raw material but indeed fully-processed material (to which we are in debt to Donald Knuth, Leslie Lamport and LaTeX) - only for libraries to have to buy it back again at exorbitant prices? Off hand, I know of no other sphere of human economic activity where this kind of thing takes place.
The fundamental idea of the template is that users must feel ownership of the data that goes into it. It's not a static box, a dump where data pulled from various sources ends up. Don't agree that Journal X allows self-archiving? Change it. Far better, say something on the text of the journal's page, and also change the template. Don't agree that Journal X is non-profit? Change it.
Unfortunately, at the moment much of the template is filled with "static" data like Eigenfactor, etc. But it should be a place where Eureka "dynamic data" is also stored. That's the whole point of doing things on a smaller scale. Eureka is different from a journal info site like Journal Info or even a journal price site like Journal Prices. We are doing things on a smaller scale, within the mathematical community (and the physics community); this allows us to take ownership of the way in which we think about journals. That's why we can give very specific subject areas like "Homotopy theory" to a journal's page. And these things are all editable... it changes as users see fit.
We could have just made the template a Big Black Box:
{{Journal Info Template}}
But we didn't. Wiki contributors must feel like they are at a wiki. Everything is changeable, nothing is sacred.
How it currently works
To use this template on a journal page, insert the following lines at the top of the entry for that journal:
{{Journal Info Template
| field1 = ''something''
...
}}
For example, the following code,
{{Journal Info Template
| Name=Antarctica Journal of Mathematics
| Image=Image:peng1.jpg
| Published by = Springer
| ISSN = 123897932
| First published = 1992
| Subject1 = Topology | Subject2 = Combinatorics
| Subject3 = Algebraic geometry
| Editor-in-chief1 = B. Simpson
| Editor-in-chief2 = M. Simpson
| Managing editor1 = M. Kapustin
| Managing editor2 = E. Black
| Eigenfactor = 0.83
| Article Influence = 1.54
| Price per article := 43.22
| Price per citation = 5.73
| Relative Cost Index = 3.43
| Composite Price Index = 1.34
| AMS 2006 Pages = 438 | AMS 2006 Price = 432
| AMS 2005 Pages = 658 | AMS 2005 Price = 412
...
}}
produces the box on the right hand side of the screen.
All of the parameters are optional, except Name. Currently, up to 5 Editor-in-chief's can be listed, and up to 10 Managing editors can be listed.
How the data is stored
We are using a bold mix of the ordinary Mediawiki system of categories, and the more advanced extension called Semantic Mediawiki. Consider the entry "Published by : Springer". When the Journal Info Template is expanded, it will contain the text
[[Published by::Springer]]
This is Semantic Mediawiki markup, to express the fact that the relation of this journal's page to the page "Springer" is "Published by". This data is stored inside Eureka.
On the other hand, we also have a category " [[Category:Published by Springer]] ". This text should be written at the bottom of the page of each journal published by Springer. Clicking on the magnifying glass in the template then takes you to Category:Published by Springer.
In Semantic Mediawiki (at least, in our oldish version), information is either entered as relations (expressing the relation of one page to another page), or as attributes, eg.
[[Price per citation:=3.21]]
In this case, the number "3.21" is not a separate page - it's just a number. But at least it's stored in the database.
Browsing the semantic data
The Semantic Mediawiki data can be accessed in two ways. The first way is to go to the Special:Specialpages section of the toolbox. Listed amongst the special pages you will find Special:Attributes, Special:Relations, Special:SearchTriple Simple Semantic Search and Types. This is all semantic stuff. Some simple searches on the data can be entered at Special:SearchTriple Simple Semantic Search.
The second way is to insert inline queries into an article. For instance, the following query asks Eureka for all the mathematics journals with an AMS survey 2006 price greater than 200 dollars:
<ask> [[Category:Mathematics journals]] [[AMS 2006 Price :=>200]] </ask>
It produces:
Algebraic and Geometric Topology, Antarctica Journal of Mathematics, Experimental Mathematics, Geometry and Topology, K-Theory, and Pacific Journal of Mathematics
When we upgrade our Mediawiki installation to a more current version, this semantic data will become much more powerful. Far more complex searches will be possible; the data will automatically be shown in a sortable table; and so on.
Explanation of the fields in the template
All the fields except Name are optional and can be left out.
- Name : The full name of the journal - this should equal the pagename (so perhaps is redundant). It is stored semantically as an attribute [[Name:=Antarctica Journal of Mathematics]] .
- Published by : Who publishes the journal. Don't forget to add the text [[Category:Published by X]] to the bottom of the journal page. Clicking on the magnifying glass takes you to the list of journals published by X; clicking on the publisher's name X takes you to the page for X. This information is also stored semantically as a relation [[Published by::X]] .
- First published : . The year it was first published. It is stored semantically as an attribute eg. [[First published:=1992]] .
- Subjects :. This is a list of up to 3 smaller subject areas that the journal is chiefly about. Things like "Algebraic topology", or "Number theory" in mathematics, or "High energy physics" or "Solid state physics" in physics (not things like "Mathematics" or "Physics" - these are stored at the bottom of the journal page as [[Category:Mathematics journals]] or [[Category:Physics journals]] ). The subject areas are stored semantically as a relation, eg. [[Subjects::Algebraic topology]] and [[Subjects::Number theory]] (note that one page can be 'related' to many other pages, and many pages can be 'related' to a specific page).
- Editor-in-chief(s) : A list of up to 5 editor-in-chiefs. Here we define "editor-in-chief" to mean the highest-level editors. This could be just one person, or it could be a small number. These names are stored in initial surname format, e.g. A. Einstein and P.A.M. Dirac (see below). This data is stored semantically as [[Editor-in-chief::Y. Eti]] , linking to the editors page.
- Managing editors : A list of up to 10 managing editors. A "managing editor" is defined to be that class of editors who are not quite editor-in-chief's, but are more important than merely "editors". These names are stored in initial surname format, e.g. A. Einstein and P.A.M. Dirac (see below). When we upgrade our system, it will be possible to enter these editors as a comma-separated list, instead of our current clumsy {{Editor1=... Editor2=...}} system. Moreover, we'll be able to store the editors as a collapsible table if there are too many.
- Eigenfactor : . The eigenfactor of the journal, as calculated at [1]. It is stored semantically as a float attribute, eg. [[Eigenfactor:=1.34]] . We will soon be able to link directly to advanced versions of this data at the eigenfactor website. We might one day obtain the eigenfactor via a live query; although this might not be the best solution.
- Article Influence : Same.
- Price per citation -> Composite price index : This information comes from www.journalprices.com.
- AMS data : The AMS has conducted - and continues to conduct - a journal price survey of mathematics journals. There is data going back to 1994, storing the number of pages and the price of the printed version of each journal. We are currently looking for volunteers to turn this data into charts. In any event, this data looks clunky the way it is currently entered. When we upgrade, we'll be able to enter it as a comma-separated list, i.e. AMS Data = $232, 232pgs, 243$, 544pgs, ... instead of the clumsy AMS 2006 Pages = ..., AMS 2006 Price = ... and so on.
How can I change the way the template looks?
The template uses CSS for formatting. The style declarations are found in the Mediawiki:Monobook.css page (click on View Source to see it better displayed).
This was perhaps not a brilliant idea, it means only sysops can adjust the CSS styling. However, each user has his/her own stylesheet file, which can be found in eg. [[User:A. Einstein/Monobook.css]] . To adjust the way the template looks (the Lord knows it does need adjusting!), I would suggest overriding the CSS declarations in the Monobook.css file by declaring new style information in your user css file. See how it looks, and if you get it looking good, or have new ideas, add them to the discussion page and we'll change the main css.
A convenient way to experiment with the CSS, if you have Firefox, is to install the Web Developer extension. This way you'll be able to edit the CSS "in real time" and see changes immediately. If you do this, you'll find the user CSS in the second-from-right-tab titled "index.php". If you're using Web Developer though, you might as well adjust the site-wide CSS, which is the third-from-the-right "index.php" tab.
How are currency values stored in Eureka?
At the moment, they're just stored as floats, see eg. Attribute:AMS_2006_Price. When we upgrade, we'll be able to institute a more advanced system; but nevertheless our current system is still workable... and here's how it works.
The basic currency of Eureka is US dollars. Thus the AMS price data always shows up with $ prefixed. If you want to enter a different currency, say for the AMS 2006 Price, just write
AMS 2006 Price = 43.22 (£21.32)
This is then stored semantically a (number, comment) pair, (43.22, (£21.32)). So eg. asking an inline query
<ask> [[Category:Mathematics journals]] [[AMS 2006 Price:=>40]]
will still give a "hit" for your page.
That's how it works at the moment. When we upgrade we'll be able to change the system so you could rather enter the natural currency first, eg.
AMS 2006 Price = £21.32 ($43.22)
This would then be stored internally as (43.22, (£21.32)), but it would be displayed as £21.32 ($43.22).
What if two editors' names are the same?
Suppose A. Einstein already edits the Journal of Marvelous Mathematics, and we are trying to enter a (different) A. Einstein as the editor of the Journal of Fishy Things. This is handled via the wikipedia "disambiguation" system as follows.
When the wiki-contributor enters "A. Einstein" in the page for the Journal of Fishy Things, he or she should click Preview and look at the journal template displayed. If the text for A. Einstein links to an already-existing Eureka page(i.e. it is not red), then the wiki-contributor should go to A. Einstein's page and ensure this is the right A. Einstein.
If it is a different A. Einstein, the wiki-contributor proceeds as follows. He or she then enters, in the journal info template box on the page for the Journal of Fishy Things, the text (for example)
Managing editor1 = A. Einstein (disambiguation text)
The disambiguation text refers to some text which will uniquely define this A. Einstein. For instance, A. Einstein (topologist) or A. Einstein (Greek mathematician). The choice of text is completely optional, it works just as at wikipedia. When the journal info template sees this text, it will still display A. Einstein in the journal info template box, but it will link to the correct A. Einstein.
Discussion
- Can we save me alot of energy (and avoid typos) by getting the numerical data directly from a suitably delimited file? --- CH 23:24, 6 September 2007 (EDT)
- Yes, it is our intention to do that. Blake was just "testing things out, see how they go". There are many things I'd like to do better 'from the start' - I'd like to have a more up-to-date version of Mediawiki for one thing, which will open up a whle new vista of functionality. But that's not the way wikis work... they appear to work by incremental improvement. This is not very efficient, but it might indeed be the most effective. For one thing, it allows everyone to feel that they are involved, which wouldn't be the case if there were a few Evil Geniuses behind the scenes laying out all the data. Did you know that Wikipedia didn't have template-functionality in its first few years? That seems incredible now, but it's true.
- Anyhow, the idea is that this higgledy-piggledy model of entering data into templates does no harm, since if at some point in the future (which could be very soon) we populate many of the fields bot-wise via data mined from text files, there shouldn't be a problem... the bot can be programmed to ignore any fields it has nothing to say about (like Editors, or Subjects) and just fill in/correct those fields it does recognize. --Brucebartlett 06:13, 7 September 2007 (EDT)
- I've put the code I used to create the math-journal stubs from the AMS data and the physics-journal stubs from the journalprices.com data on my user page. The output of my parser scripts is meant to be run through pagefromfile.py in the pywikipedia bot collection. I was a little wary of trying to put the new Journal Info Template on pages which already had the AMS data, since that's a more involved bit of text manipulation. Blake Stacey 12:28, 7 September 2007 (EDT)
- Where can we discuss (by email?) my concerns about eigenfactor? C.f. my complaint that statistics with a dubious or murky interpretation are suspect. Am the only one who has studied what eigenfactor.org says about their statistic? I am worried about teleportation but can suggest alternative models which would give different statistics. BTW, the inventors of eigenfactor are physically accessible. ---CH 23:24, 6 September 2007 (EDT)
- I suggest we discuss it at Template talk:Journal Info Template/Eigenfactor. --Brucebartlett 06:13, 7 September 2007 (EDT)
- I suggest change EF and AI to percentile scores for these (45 for 45% and so on). ---CH 22:59, 6 September 2007 (EDT)
- Well, I suggest we have both. Something like Eigenfactor : 0.38 (45%), I don't know. The trouble is the percentiles currently shown at www.eigenfactor.org are with respect to the entire JCR database; at least that's what I understood from private communication with Carl Bergstrom, who is in the Eigenfactor team. They are upgrading fast though, and soon they'll be able to offer percentiles "just within mathematics journals" or "just within physics journals" which is the kind of thing we want, right?
- With Semantic Mediawiki we can also calculate these things ourselves. We could for example display the ranking, like "Eigenfactor : 0.38 (52nd)" which would mean "it's 52nd in the list of Eigenfactors". Unfortunately this kind of functionality requires the latest version of SMW, which in turn requires PHP5, which we'll upgrade to soon. --Brucebartlett 06:13, 7 September 2007 (EDT)
- Classical Quant Grav (should be italicized automatically by template), i.e. the standard abbeviation, should appear instead of picture (what could possibly go there?) under "Classical and Quantum Gravity". ---CH 22:59, 6 September 2007 (EDT)
- Well, I like to think that the picture of the front cover of the journal can go there, like so:
- But I wouldn't want people to just pull these images off the webpages (which Springer might not find to be fair use, though it possibly is) without permission. Perhaps they could take a digital photo of the front cover? As for abbreviations, I don't think its necessary to display this anywhere; it seems like the kind of information which is easily derivable from the full name. --Brucebartlett 06:13, 7 September 2007 (EDT)
- Add field Profit? (Y/N) ---CH 22:59, 6 September 2007 (EDT)
- Yes, definitely! We should just work out a good system for this classification. I guess by "Profit" you mean "this journal is sold for profit". Or do you mean "this journal is published by a non non-profit organization". --Brucebartlett 06:13, 7 September 2007 (EDT)
- Add field arXiv? (Y/N) meaning: can authors archive their own eprints? ---CH 22:59, 6 September 2007 (EDT)
- Yes, definitely! I was a trifle worried that the SHERPA data wasn't quite accurate for our little field of mathematics... the point is we really want wiki contributors to enter this data themselves, rather than us blindly taking the word of some big database like SHERPA. But perhaps we should start off that way.
- Add options for ISSN-print and ISSN-online. ---CH 22:59, 6 September 2007 (EDT)
- Yes, you're right. I wasn't even aware that these two different numbers existed. --Brucebartlett 06:13, 7 September 2007 (EDT)
- Use easily guessed abbreviations in the template so more information is visible in a given amount of vertical space, e.g. "EF %", not "Eigenfactor percentile". (The talk page for the template should of course explain these abbreviations.) ---CH 22:59, 6 September 2007 (EDT)
- Well, perhaps, but I'd like to keep things self-explanatory as far as possible. A first-time wiki editor should be able to adjust some numbers or fields on a journal's page without having to look at the documentation.
- Having said that, some of the fields take up way too much space at the moment and are handled in a clumsy way. For instance, to add the editors you need to type "Editor-in-chief1 = A" and "Editor-in-chief2 = B", and so on. The same goes for the clunky AMS data. These are better entered as a comma-separated list, eg. "Editor-in-chief = A, B, C, ...". Unfortunately that requires the Winter extension, which requires we upgrade to PHP5, which requires... But we'll get there.
- Why not write the template so the journal title is also an external link to the home page? ---CH 22:59, 6 September 2007 (EDT)
- Good idea. I'll try and do that. --Brucebartlett 06:13, 7 September 2007 (EDT)
- Did I misunderstand the point of Semantic Wiki; I thought the point was to link to the eigenfactor.org page for this journal if you click on the magnifying glass labeled EF? Simply listing in "order" all instances where "eigenfactor" appears at Eureka doesn't seem very useful to me. Compare the ability at Eigenfactor to order physics journals (say) in order of decreasing Eigenfactor percentile. ---CH 22:59, 6 September 2007 (EDT)
- We can have both - we can link to the eigenfactor page (like this link) and to our internal Eureka list of articles. However, neither option is functional yet. Linking to Eigenfactor's journal pages is not quite possible yet due to the fact that they use strange abbreviations for journals as a parameter in the http address.
- You're right - it is pointless for the magnifying glass to just link to the Special Semantic Search page at Eureka and simply list all journals which have the field "Eigenfactor". As I said in the documentation, when we upgrade to SMW 0.7, we'll be able to do far more advanced things with this, like automatically sorting it, displaying it any way we want, and with any queries we want. For now it's just a stop-gap so that people can get ideas of the 'potential'. --Brucebartlett 06:13, 7 September 2007 (EDT)
