11 Saturday Mar 2017
Posted in Spotted in the news
The open source monthly Web crawl Common Crawl now “contains 3.08 billion+ web pages” and 250TB of uncompressed data. Back in September 2015, when this blog last took a look at Common Crawl, that was 145TB.
21 Tuesday Feb 2017
Posted in Spotted in the news
European Space Agency announces new open access policy for images. Now CC BY-SA.
Space in Images (20,000).
Space in Videos (3,200).
Jorge Manes Rubio, ESA Advanced Concepts team, temple/telescope for future lunar colony on the Moon.
15 Wednesday Feb 2017
Posted in Spotted in the news
Social Media Research Toolkit is an up-to-date and academic list of free tools for bulk-collecting and investigating various forms of social media.
15 Wednesday Feb 2017
Posted in Spotted in the news
TagWalk: The Fashion Search Engine is a very nice high-end ‘trends search’ experience for the fashion industry. It kind of feels like you’re using Anna Wintour’s personal search-engine, but it’s public. Extensively tagged by hand, by a small team of ‘industry insider’ curators. It’s fast too. Humdrum academic archive discovery searches could learn a thing or two here, in terms of slick navigation and speed.
The only thing I’d change would be the ability to switch through from womenswear to menswear, so that when you’re looking at Spring/Summer 2017 “tag” in womenswear results, you can instantly flick over to the same “tag” in menswear for visual comparison. I’d also add “hat” and “cap” to the tagging, since they’re currently missing, though that may be because designers are not currently sending hats down the catwalk. At the back is “moodboards”, which I didn’t explore, but I presume you can pull your own custom moodboards of designs.
08 Wednesday Feb 2017
Posted in Spotted in the news
The image search at creativecommons.org has been given a makeover. The beta of the new design is at a new URL of ccsearch.creativecommons.org, rather than the long-standing search.creativecommons.org.
07 Tuesday Feb 2017
Posted in Spotted in the news
The Metropolitan Museum of Art has just placed 200,000 CC0 images of its art and others items in its collection. So far the site is holding up nicely in terms of managing the traffic.
The test files I had from it were large, at around 4000px, and were also crisply photographed. The site’s search works cleanly and a single click downloads the largest file size as a .jpg. The only thing I’d change is to allow a Web browser’s Back button to work with the site’s search results.
06 Monday Feb 2017
Newly announced for the UK…
“Today Jisc announced that OCLC, the global library cooperative, has been awarded the contract to develop a new national bibliographic knowledgebase (NBK).”
Judging by the initial press-release, the focus seems likely to rest first on cohering UK academia’s metadata management for digital book collections. This will in time…
“enable shared bibliographic metadata to flow into … global search engines”
Hopefully that means Google Search, as well as Google Scholar (which are two separate systems and databases).
05 Sunday Feb 2017
Posted in Spotted in the news
Kapersky Labs is developing the FFForget tool, aiming to create a local backup of your Facebook. Hopeful it will also archive the Facebook Groups you moderate, while allowing local search of their full-text.
“The service is planned to go live in 2017 based on user’s interest. Subscribe now for free and become one of its first users!”
05 Sunday Feb 2017
Posted in Spotted in the news
From the CORE blog…
“CORE is thrilled to announce that it currently provides 5 million open access full-text papers.”
A very worthy achievement, given the surprising difficulty of auto-harvesting full-text from university repositories. I’m happy to say the CORE full-text has long shown up in JURN’s search results.
03 Friday Feb 2017
Posted in Academic search, Spotted in the news
Lecture Search at findlectures.com is a new search tool, or at least it’s new to me. It appears to have launched in the summer of 2016, which is probably why I missed its launch. In the UK it would have been drowned out by the news of our glorious Brexit.
Lecture Search aims to find ‘intelligent talk’ files such as conference and academic lectures, and it does what it claims. A few early observations:
* Seems to be running from a hand-curated URL list. There’s evidence in the results that the last indexing run may have been in early 2016?
* Includes YouTube and Vimeo as sources but seems to have a filter on them, presumably via indexing only selected channels.
* Searchers should use NOT keyword rather than -keyword to knock out search words from results.
* Nice range of limiting facets, in the sidebar.
* One annoying pop-up nag-box, but it was easily killed with AdBlock Plus’s “select an element to hide…”.
* Relevance ranking is definitely not Google-licious, as it the case for all such Summon-like services. For instance: search for “cave art”, get “The Complete Poetry of Cesar Vallejo” as the first result. That page’s text happens to mention Vallejo once did some research on “cave art”, but then presumably the prestige of the result’s loc.gov URL lifts the result up to No.1.
* Not indexing the BBC’s hundreds of In Our Time .mp3 podcasts, which seems a pity.