RPI, Dropbox, Mapping, More: Sunday Afternoon Buzz, March 2, 2014

Rensselaer Polytechnic Institute has digitized its newspaper archives. “Issues of The Poly from 1885 to 1976 can be searched online by anyone. An earlier iteration of The Poly that only lasted for one semester, the fall of 1869, is also available in the archives. Because some syndicated content is still protected by copyright, issues of The Poly from 1977 to 2001 can only be accessed on campus or with a current Rensselaer login.”

Nifty. Mapillary is trying to build a crowdsourced version of Google Street View. “With the app, a user can choose to collect images by walking, biking, or driving. Once you press a virtual shutter button within the app, it takes a photo every two seconds, until you press the button again. You can then upload the images to Mapillary’s service via Wi-Fi, where each photo’s location is noted through its GPS tag. Computer-vision software compares each photo with others that are within a radius of about 100 meters, searching for matching image features so it can find the geometric relationship between the photos. It then places those images properly on the map, and stitches them all together.”

Because all work and no play: Marvel is making thousands of comics available via an iOS and Android app.

You can now have unlimited secret boards on Pinterest.

Google is now showing restaurant menus in its search results, but this is a US-only feature at the moment…

Nice! Over 20 ideas and resources for teaching with Instagram. I think I found out about this from Pat O. Thanks Pat O!

The Economist has an interesting article about Wikipedia and the challenges it faces as it evolves.

From Lifehacker: 7 Downloads and Extensions to Make Dropbox Even More Awesome. Good afternoon, Internet…

I love your comments, I love your site suggestions, and I love you. Feel free to comment on the blog, or @ResearchBuzz on Twitter. Thanks!

Vermont, DMOZ, Wikipedia, More: Tuesday Buzz, February 4, 2014

Thanks to PIBuzz for the pointer to VTDigger’s announcement that it has launched a searchable campaign finance database. “The database features complete lists of donations to candidates for Vermont House, Vermont Senate and statewide seats for the 2011-2012 and 2013-2014 election cycles. It offers information about how much candidates have raised and how much donors have contributed across political campaigns.”

The state of Mississippi has a new Web site for job seekers.

Tumblr has added support for SSL but, for some reason, does not turn it on by default.

Pinterest has updated its acceptable use policy and is cracking down on paid pins.

DMOZ has apparently just dropped a million sites according to a post on SEO Roundtable. Does anybody use DMOZ anymore?

The White House has announced its first Maker Faire. What, so the NSA can teach you to bug your own computer?

The Google Cast SDK is now available. “The Google Cast SDK is simple to integrate because there’s no need to write a new app. Just incorporate the SDK into your existing mobile and web apps to bring your content to the TV. You are in control of how and when you develop and publish your cast-ready apps through the Google Cast developer console. The SDK is available on Android and iOS as well as on Chrome through the Google Cast browser extension.”

Want to learn how to write Wikipedia articles? Apparently there’s a course for that. (And it’s free.) “The course covers the technical skills needed to edit articles, and also offers practical insights into the site’s collaborative norms and social dynamics. Students graduate with a sophisticated understanding of how to use Wikipedia both as a reader and as an active participant.”

Transparency reports are getting even more transparent. “The data from Google shows a significant growth in internet content collection from its products by the NSA. In the first six months of 2009, the company gave the government data from up to 2,999 customer accounts, a figure that grew to between 12,000 and 12,999 customer accounts by the second half of 2012 before dipping to under 10,000 accounts in the first half of 2013.”

Apparently it’s a dud: Google ordered to move “mystery” barge. And now I feel compelled to explain that that was a “Mystery Date” joke. All will be revealed at http://en.wikipedia.org/wiki/Mystery_Date_(game) . Good morning, Internet…

I love your comments, I love your site suggestions, and I love you. Feel free to comment on the blog, or @ResearchBuzz on Twitter. Thanks!

Wikipedia Trying a Different Editing Approach

If you don’t use Wikipedia much, you might think that it’s kept “open” all the time, and that any page may be edited at any time by any person. That’s not correct. Sometimes pages have to be locked against editing, usually because they have become a target due to something happening in the news. Wikipedia says only about 0.1 percent of the 3.3 million articles on the English Wikipedia are actually being protected against editing.

Wikipedia announced last week that it is taking a different tack in protecting pages. For the next couple of months the site site will try something called “Pending Changes”. (The Wikipedia page for pending changes is http://en.wikipedia.org/wiki/Help:Pending_changes.

Normally when pages are protected there’s a little lock in the upper-right corner. (For examples of semi-protected pages see here and here.) These pages can’t be edited anonymously or by newer Wikipedia accounts; see more details on semiprotected pages here.

But with these new pages you’ll see a magnifying glass. The page will still be open to edits, but the edits will have to be reviewed before they’re allowed into the article. Anyone will be able to see proposed edits by clicking on a “Pending Changes” link, but of course not everyone will be able to approve the edit.

Apparently there’s a 2,000 page maximum for this trial, which might be why it was so tough for me to find an example of a Pending Pages page. I finally had to use the example in Wikipedia’s blog, the Solar System Wikipedia page. You can still see the history, but the magnifying glass makes it clear when the last edits took place.

This trial is interesting enough, but I also found some other icons in the upper-right corner near the magnifying glass or lock interesting. Icons denote Featured Articles and Good Articles in addition to any protection the article has on it. (Fun tool: click this link for a random Featured Article.) Some articles also have a icon showing they’re available in a spoken word version.

Answers.com Pushes Answers to Twitter

Answers.com, which you might remember from WikiAnswers or ReferenceAnswers, announced last Wednesday a new alpha feature called “Hoopoe,” which ties in with Twitter. Hoopoe’s Twitter account is at http://twitter.com/answersdotcom.

Here’s how it works. Either send a tweet to @AnswersDotCom or write a tweet with the hashtag #AnswersDotcom or #hoopoe . Answers.com will send an autoresponse with a snippet of information and a pointer to the answer on its Web site.

The first thing I did was send a tweet to @AnswersDotCom asking “What is a hoopoe?” No more than a minute or so later I had the tweet you see above. The URL ponted me to an Answers.com page that provided information from a dictionary, Columbia Encyclopedia, Western Bird Guide, and a huge Wikipedia article with lots of images. I am now very confident about my basic hoopoe knowledge.

Next I tried a more abstract question, “When is the sun coming up tomorrow?” I tried that twice with #AnswersDotCom and #hoopoe. Be careful about including extraneous text in your question; Hoopoe will try to answer all of it. (It would be great if it only tried to answer the text set before the hashtag. That way you can warn all your followers you’re playing footsie with an autoresponder.) AnswersDotCom couldn’t answer the question, referring me to the site instead.

Thinking that perhaps that question didn’t provide enough data (like where I was) I asked a simpler reference question, “What is the square root of 12?” I got an answer from Answers.com but it was incorrect. And it completely ignored my similar question from a few minutes later — I think it searched for “Square root of” and gave me the first thing it could find in its database.

So if you’re looking for reference-type information, Answers.com’s Twitter service is fast and good. And #hoopoe is a hashtag I could actually type on my cell phone without too much trouble. On the other hand if you want math questions or almanac questions answered, you’ll have to keep waiting for Wolfram|Alpha to come out with a twitter service.

Making A Reference Page From Several Resources: Mashpedia

Over at MakeUseOf I read about a recently-launched resource called
Mashpedia. Though it might sound like a wiki for potato recipes, it is actually a cool tool from gathering reference information from several places into one easily-scanned page. It lives at http://mashpedia.com/.

The front page shows you some hot searches but you can also do a keyword search. This is topical searching, so think directory-level searches, not anything complicated like you might do for a full-text search engine. I did a search for Ben Franklin.

The resulting page pulled together a variety of information for me, including Wikipedia details, videos from YouTube, two tabs’ worth of content from Digg, a Twitter feed, photos from Flickr, what looks like feeds from Google News and Google Images, and even a book search result. Every set of information resides in its own collapsible section; you can also set sections to show only one result or to show all results (the “all results” actually paginate so you’re not going to get a giant scrolling page.)

Now, the search for Ben Franklin had mixed results. On the other hand, it started off with a great Wikipedia article, the quotes pages in the Web results were good, his autobiography popped up in the Google Books section, and there were actually a lot of relevant YouTube and Flickr results. On the other hand the name “Benjamin Franklin” has bene used for so many things that there was a lot of gunk in the results as well. I tried a second Mashpedia search for something less ambiguous — Butler University (congratulations on your great NCAA basketball season.) These results were much more focused, except the Digg results were a little odd (looked like the words were being searched for separately.)

This would make a handy quick reference. It could be even handier if some of the searches could be focused more for ambiguous terms (it might be tough, though.) I also wonder if some of Yahoo’s Web services wouldn’t be a wonderful match for the existing offerings here? I’m thinking specifically of the Term Extraction Service.