Talk:Google Scholar

Not as good as described in this article

It seems that google scholar is not as good as described in this article. A perfect paper on the subject matter is written here You can check it out for yourself, what is written is really true. Would it thus be possible to include a milder version of my critic on this wikipedia page, like it exists in the German version of this article, without having it deleted? I thank you in advance. —The preceding unsigned comment was added by 194.39.218.10 (talk • contribs) .

The German article referred to is aat http://de.wikipedia.org/wiki/Google_Scholar

would 194.39.218.10 please give some contact information? ---

I found that your article was somewhat informative. However, it seemed to be highly biased against Google Scholar by blasting it at every opportunity. An example: "The "Googlemania" fueled by the enormous media publicity and laymen's ecstasy rubs off on Google Scholar and makes otherwise learned people disregard reality." —The preceding unsigned comment was added by 131.202.42.70 (talk • contribs) .

Not as bad either

GS is a complex project that has changed direction several times. It's history and current state needs to be described at kept up to date, and the BETA warning on the page is certainly appropriate.

• It is different in one conspicuous way from almost all other services--it is at no cost to the user. -- theres a nother prominent one, with an aticle stub, Scirus.
• GS also has the special feature that in its current incarnation, it can be used as a citation index. There are a number of links to be made here. The German version does discuss this.
• It was originally aimed at linking to open access versions. Publishers began adding crawlable links to their subscription versions, now almost universal for major scholarly journals. At present, it finds the links to the subscription versions only, and links to a search of the main Google space for open access versions.

What it will be like a month from now, I can not guess, because it does not depend only on Google but upon its sources as well.

• Wikipedia should be good at dealing with all this, if those who use it keep updating the article.
• NPOV is not just wording--it is (or, imho, should be) describing what is good and what less good.
• "references" here are a little tricky, because everything published is out of date,
  • • including Jacso, which is now two years old and deserves a warning.
    • similarly the item from UBC is a single minor pragraph.
    • and JR's article is mostly about A&I resources in general ,and does not even mentiont GS.
    • the Nature links are hopeless out of date.
    • but the HEP link is both good and recent.
• I think the least controversal way to at least hint at this is to add the dates to the references that do not already have them, and a word of comment. I am about to do so.
  • I defer removing the JR and the UBC to see if there are protests;
  • There are more up to date ones, and I will add them, but not today. see Google: bakkalbasi google scholar
• As in many computer oriented subjects, I think WP is likely to be& remain the best source, because we can keep it up to date in whatever detail necessary.

DGG 19:29, 12 October 2006 (UTC)

bloody useless

Bloody useless if the work it's pointing at that you're trying to read is on JSTOR or suchlike, which requires a password (bugmenot doesn't work) to use it. Thanks for the easy access to.......articles I can't access... Thanks Google! -Tim Rhymeless (Er...let's shimmy) 08:55, 12 October 2006 (UTC)

but JSTOR will reveal the authors, who you can contact to ask for a PDF or reprint Jerry Vanclay 04:25, 12 January 2007 (UTC)

I have removed the paylink to the $30 JSTOR article which you were probably talking about. Such links are not admissible per Wikipedia:External links#Sites requiring registration (see also "Links to sites that primarily exist to sell products or services" on the same page). If such a publication needs to be cited as a reference for information in the article, standard bibliographical information should be used to identify it - just like a book should be referred by ISBN etc., not by a link to a web shop which sells it. Regards, High on a tree (talk) 23:49, 21 February 2008 (UTC)

Perhaps it would be worth mentioning in the current article that the majority of use of Google Scholar comes from academia, specifically institutions that use it as a tool with their ezproxy servers. That's why you don't have an [[open access] search-modifier, although you can always narrow down the search results via filetype:PDF and other tricks. -- kanzure (talk) 18:26, 29 April 2008 (UTC)

but help is on the way

It depends where you are. if you are outside a major university, your comment is certainly not without foundation, though in general it is possible for anyone to --eventually--get copies of articles that are mentioned through Interlibrary Loan; this can be a rapid process, but usually is not. If critical, it is possible to purchase most individual articles from the publisher, at a cost ranging from $5 to $35 an article, or through document dellivery services such as the British Library which offer copies of almost anything through their {http://direct.bl.uk/ copy service} at similar prices.

For a discussion of ways to solve the fundamental problem and provide everyone with immediate access to all published research articles, see open access. I and many others have been engaged in advocating for this for a number of years, and there are modest signs of progress.

Even outside a university it can be useful:

• Some journals are already Open Access, including such major journals as PLoS Biology
• Many journals become available by open access after a an Embargoperiod.
• Some individual articles are now available.
• It lists libraries that hold the book or journal, so you can visit them
• It links to later and related articles, which may be available
• but, most important, it provides a good link to Google, which lists primarily versions that are available without charge. In general, the link to Google is a well-constructed one, which should find exactly what you want, probably more precisey than a search made by an individual user.

Unfortunately, there is no direct way of finding out if you will have access to an article except by trying all the possibilities. GS is useful in at least bringing them together.

you should also try Scirus, which can be set to give free web copies only.

If you are searching for recent biomedical articles, you may do best at PubMed, which will often give listing to a free copy available through PubMed Central

---

(I intend this as part of an article in WP about how to access journal articles, and would like suggestions for an article name. Most of it is not specific to GS, which may just have a link. I do not necessarily want to add it to the interlibrary loan page, which is general information and what I've just written is subject to very rapid change.) DGG 19:29, 12 October 2006 (UTC)

Contradiction

I've just removed the following from the article:

Using its "group of" feature, it shows the various available links to the journal article. If the article is available freely on the web, those links will be shown, but otherwise it provides only links to the subscription-based sites (At least, that's what is was doing in September 2006. In some earlier versions, its "Group of" feature sometimes also provided a link to free full text versions of articles.) Access to non-subscription versions is now provided by a link to Google, where one can find such open access articles.

(emphases added)

As you can see, this paragraph contradicts itself (and jumps around in time to boot). I can't make sense of it; can anyone else? --zenohockey 19:28, 2 December 2006 (UTC)

It is widely known and cited as listing non-subscription versions. It once did, it no longer does, and this seem ot in the least confusing of contradictory. I have reworded. Please talk before you revert.DGG 03:24, 3 December 2006 (UTC)

Scientists vs. scholars

I've changed 'scientists' to 'scholars' in the first paragraph, in order to more accurately reflect the coverage of GS. I've also added a brief gloss on the phrase 'stand on the shoulders of giants'. fi99ig 21:18, 1 January 2007 (UTC)

Lulu?

Any idea whether it extends it's dbase to lulu.com and other releasing mediums of that nature? I've found a lot of great thesis published there in the past and am unsure whether to search there as well as the usual places for academic articles or whether scholar.google covers them all? Jachin 08:02, 11 June 2007 (UTC)

Serious problems with Google Scholar "cited by" option

I like Scholar. It is so much easier and faster than Web of Science that I use it much more regularly, even though my university pays for WoS. But I have been bitten by it so many times now that I spent the larger part of the day performing the following little survey. I chose a landmark article and counted the number of references citing it, according to GS and according to WoS. Here are the results:

• 22 papers cite this article and appear in both GS and WoS.
• 29 papers cite this article and appear only in WoS.
• 2 papers cite this article and appear only in GS.
• 5 conference papers cite the article and appear only in GS (WoS specifically does not index conference papers).
• 6 additional items listed by GS are not really papers; they include 3 dissertations, 1 book review and 2 unpublished working papers.

Conclusion: Google Scholar may be convenient and fast, but under no circumstances should it be trusted to give you a full list of citing papers. Interestingly, most of the papers which GS does not list are found if you search directly for them in GS; they just don't appear as works citing the original article. --Zvika 11:50, 20 August 2007 (UTC)

GS is not reliable for older articles in technology, nor are any internet based sources. This paper was published in 1969. DGG (talk) 04:40, 22 August 2007 (UTC)

I'm not really sure what you mean. Web of Science is an internet based source too, and it seems to work fine. Also, almost all of the papers listed by WoS can be found online and are available through GS by searching for them directly; they just aren't listed as citing the original paper. It's clearly a bug, or an inaccuracy in GS's OCR. --Zvika 06:29, 22 August 2007 (UTC)

You are correct that I oversimplified., but it was the end of the day.

In the sense I mean it,internet based means deriving its fundamental information from the internet.

WoS derives the information about papers and references from examining the references in papers and coding them. Most of these of course are now available in electronic form from the publishers, though not all. and it includes the years for which electronic versions have not yet been made available. It essentially includes all publishers--it tries to get publishers to donate the access, and so great is the prestige from being included that essentially all publishers do; for those that do not, it simply buys a subscription. One then accesses it through the internet (in the past, via print or CD, options which remain available today). GS goes from web links between papers as available on the internet from the publishers, supplemented by other internet source from Google and google books. It has no mechanism at all for accessing print material outside of Google books. It also is dependent on the goodwill of the publishers. Until very recently, neither Elsevier nor American Chemical Society made current material available to it. This is no longer the case, except that not all the material from these publishers has in fact been entered yet. So references to material there made from other journals is present; references from their own journals to anything has not been present. I have not checked on how complete it is since early July, & it changes fast.

You are right that the time depth in GS is increased, now that backfiles are available for most journals. Again, I have not checked since early in the summer., and I should recheck. But the exact methods GS uses are not made public, and one must get information through reverse engineering.

So I would like to do a failure analysis of your results, and I'd appreciate if you could post them here, or email them to me through my WP link, or I'll send you my email address. I want to see if I can explain the results. If not, I want to consult my colleagues, included others studying this source,and those on the appropriate professional list. WoS at least replies to inquiries. It's not a bug in the usual sense, and I don't think GS does OCR (GB does, of course, and the errors there are very many). If you know of GS doing OCR or other material I'd love to know about it & so would other science librarians. DGG (talk) 17:50, 22 August 2007 (UTC)

The list is quite long and I'm pretty sure it doesn't belong here, so I emailed it to you. If anyone else is interested, let me know. Regarding the source of the information, as I said, almost all of the papers that GS doesn't identify can be found on the internet, and can even be found using GS if you search for them directly.

I don't know if GS does OCR or just uses OCR results implanted in PDF files, but I've seen it find things resulting from incorrect OCR identification many times. --Zvika 11:04, 23 August 2007 (UTC)

Citation (example) for erroneous additional search results

Searching for "autor:witten-e +arxiv" leads on result page 45 to the result

Edge excitations of the Chern Simons matrix theory for the FQHE
ID Rodriguez, J Kluson, RG Cai, B Hu, Y Zhang, R … - eprint arXiv: 0812.4531, 2008 - adsabs.harvard.edu
... Authors: Rodriguez, Ivan D. Publication: eprint arXiv:0812.4531. Publication Date:
12/2008. Origin: ARXIV. Keywords: High Energy Physics - Theory. ...

in which the string "Witten" does not appear at all. Direct link to search result: http://scholar.google.com/scholar?q=autor:witten-e+%2Barxiv&hl=de&lr=&start=440&sa=N direct link to resulting paper: http://arxiv.org/PS_cache/arxiv/pdf/0812/0812.4531v2.pdf 84.163.117.6 (talk) 22:40, 17 February 2009 (UTC)

Does not cover individual university pages?

This part seems to be outdated: "As of December 2006, it provides access to both published versions and on major open access repositories, but does still not cover individual university pages". Nowadays Google Scholar provides links to self-archived versions on personal home pages, etc. But do we have a reliable source for this? — Miym (talk) 09:11, 2 May 2009 (UTC)

It should read individual faculty pages, & I have made the change. I will look for a source, other than my own postings in various places on the subject, and ask an expert I know who has published on similar questions. DGG (talk) 22:08, 2 May 2009 (UTC)

Some search results are not entirely traceable....

Could experts diagnose the problem....???

See my work
http://scholar.google.com/scholar?hl=en&source=hp&q=Madhyamaka%20materialism%20permanence&um=1&ie=UTF-8&sa=N&tab=ws

http://books.google.com/books?hl=en&lr=&id=yiz9gPJBiNIC&oi=fnd&pg=PR9&dq=Madhyamaka+materialism+permanence&ots=GHRNbLohIY&sig=EyCvUng1UsgzYvSJq-ENYCdbVqs#v=onepage&q=dialectical%20analysis%20employed%20by%20the%20Madhyamaka%20is%20evident%20in&f=false --222.64.26.148 (talk) 00:55, 21 August 2009 (UTC)