24 Monday Feb 2014
Posted in New titles added to JURN
JURN is now experimentally including thesis repository records in search results, via indexing some of the newer and more discriminating global repository search services.
22 Saturday Feb 2014
Posted in Official and think-tank reports
New White Paper from commercial publisher SAGE: Discoverability of Scholarly Content: Accomplishments, Aspirations and Opportunities, written after the moment…
“in 2013, [when commercial paywall] academic content providers around the world were forced to consider alternate methods of open-web discoverability when Google’s primary web search ceased its special treatment of access-controlled scholarly materials”
The consusus here and elsewhere seems to be ‘just add metadata’ to open content, and it will slot right into existing closed discovery systems. Good luck with that one, and don’t forget to shut the stable door.
Interesting comment spotted in the paper, in relation to the oft-heard complaint that Google uses closed proprietary search algorithms…
“The research also noted the variability among [academic library Web-scale] discovery services’ proprietary search algorithms, which — lacking transparency — disallows local customization, which libraries and end users expect.”
What it doesn’t mention is the quality of the ranking done by the proprietary search algorithms in services such as Summon. To see what I mean, try the dozy search | what is history carr | in Summon, then try it again in JURN.
20 Thursday Feb 2014
Posted in Economics of Open Access, Spotted in the news
The Amazon Public Data Sets service offers free data storage for useful public domain Big Data sets, on Amazon’s uber-servers. Analysis tools too, it seems.
20 Thursday Feb 2014
Posted in Spotted in the news
The Common Crawl bots spent 2013 crawling the Web. Now their 102TB / 5 billion page index is available to anyone who wants it. For free. Re-use it freely too, on what is effectively a CC0 licence.
20 Thursday Feb 2014
Posted in Spotted in the news
Elastic Search 1.0 has just released. It’s a search-engine, open source, with four years of development under the hood. Aimed at big businesses, but free. Currently trying to swing itself around to head toward the gap that’s looming due to the shortage of human Big Data analysts. InfoWorld magazine has a straightforward overview.
20 Thursday Feb 2014
Posted in JURN's Google watch
In an unusual move Google appears to have created its own Custom Search Engine, Custom Search for K-12 Computer Science Education. For the benefit of those outside the USA, “K-12” isn’t the name of some obscure Linux module. It seems to be U.S. educational jargon indicating: “state schooling for kids aged 5 to 16”.
14 Friday Feb 2014
Posted in Spotted in the news
Web hyperlinking to freely available online content does not amount to publishing an illegal communication, says a major new ruling by the Court of Justice of the European Union. Linking to open content cannot communicate it to a new public audience, since the content is already public.
However, the ruling says nothing of the taxing of Web links to open content, a loopy idea proposed by French socialists among others.
11 Tuesday Feb 2014
Posted in Spotted in the news
The US Defense Advanced Research Projects Agency (DARPA) want to build Google-style custom search engines for spooks, or as they phrase it… “domain-specific indexing of web content [with] domain-specific search capabilities”. Might be nice if they could also invite worthy non-profits like JURN to park up on their uber-servers for free, with maybe 10,000 domain URLs to play with — compared to the 5,000 URL limit Google places on its CSEs.
06 Thursday Feb 2014
Posted in Spotted in the news
TED has the full set of statistics from Coursera, one of the leading MOOCs…
05 Wednesday Feb 2014
Posted in Ooops!
Google Search has effectively just de-indexed most of the DOAJ, due to the DOAJ’s recent switch to a new dynamic results approach which did away with 1.4m Google-friendly static HTML pages.
Just before Christmas: 1.4m pages, including individual article records…
Today: 16 pages indexed…
