On a site as big as NYTimes.com, changes and updates occur on a daily basis. Here’s a look at some of the features we’ve added in the past six months.

Inside NYTimes

You’ve probably noticed the “Inside NYTimes.com” module, which appears on the home page and across the site.

Swipe left or right on a touch-enabled device to load the previous or next item

Readers with iPads and other touch-enabled devices can now drag to scroll horizontally through the module (instead of clicking the arrows in the top right corner).

Simple Swipe Gestures

Here’s another feature for touch-enabled devices: on articles and blog posts, you can now use simple swipe gestures. Swiping left-to-right or right-to-left will load the page for the previous or next item in the current section or blog. (If you’re viewing the first item in the series, swiping will have no effect.)

Deep-Linking and Highlighting

NYTimes.com now provides anchor links or “deep links” to specific paragraphs: to the URL for an article or blog post, you can add a hash mark (#) followed by the letter p and the number of the paragraph you want to link to. For example, this URL goes directly to the fifth paragraph, #p5:

http://www.nytimes.com/2010/11/30/technology/30jumo.html#p5

In longer articles (especially in single-page view), it isn’t always easy to determine the paragraph number without manually counting from the top. Try tapping the Shift key twice: this toggles pilcrows (paragraph marks). Pilcrows help you scan the page and quickly identify paragraphs. You can also click a pilcrow to get the link to that specific paragraph (the link will appear in your browser address bar).

And there’s more! You can highlight specific paragraphs by using the letter h after the hash mark (instead of the letter p). For example, this URL highlights the fifth paragraph instead of jumping directly to it:

http://www.nytimes.com/2010/11/30/technology/30jumo.html#h5

You can even highlight multiple paragraphs. This link highlights the third and fifth:
http://www.nytimes.com/2010/11/30/technology/30jumo.html#h3h5

But what if you want to highlight a specific sentence instead of the full paragraph? You can! Just add the letter s and the number of the sentence:

http://www.nytimes.com/2010/11/30/technology/30jumo.html#h5s1

You can even highlight multiple sentences by using commas to separate the numbers:
http://www.nytimes.com/2010/11/30/technology/30jumo.html#h5s1,2

(Of course, in this example, that’s the equivalent of just #h5, but you get the idea.)

Let’s put it all together. In the article Eating the Irish, we want to:

• Go directly to the paragraph 2: p2
• Highlight paragraph 3 in full: h3
• Highlight sentence 2 of paragraph 4: h4s2
• Highlight sentences 1 and 3 in paragraph 5: h5s1,3

Deep linking and highlighting

Add it all up:

http://www.nytimes.com/2010/11/26/opinion/26krugman.html#p2h3h4s2h5s1,3

Try it out! But just one word of caution: new articles and blog posts (especially blog posts) may be updated frequently while a story is developing. So the position and content of a paragraph may change throughout the day — please bear this in mind when linking to and highlighting paragraphs in newer stories.

That’s a brief look at some recent tweaks and features. We’ll keep you posted as we continue to enhance NYTimes.com. To learn even more, come to Hack Day this Saturday. Many NYT developers will be in attendance, so feel free to ask them how they bring new features to life!

It’s been far too long since we last looked at some creative and instructive uses of Times APIs. Here’s a quick sampling of apps and tools that will help you explore Times APIs — and maybe even inspire you to build something of your own.

1. This dynamic roll-call vote map uses data from the Times Congress API.

Roll-Call Vote Map

2. Take a look at some sample VBA code for use with the New York State Legislature API.

3. Justin Blinder’s visualization of tweets was written in Processing and built with the Times Article Search API, the Twitter API, the BackTweets API and the Google Maps API.

Justin Blinder Visualization: We Read, We Tweet

4. Play New York Times Roulette: get a random article or blog post, compliments of Daniel Vydra, the Times Newswire API and Google App Engine.

Post your own ideas and creations in the comments, and don’t forget to visit the Times Developer Network and try the API Console.

Our series of posts about external uses of Times APIs continues. Today we’re highlighting the University of Puget Sound’s use of the Times Best Sellers API.

A couple of months ago, we looked at the Dallas Public Library’s use of the Times Best Sellers API. This week, we have another great example of how libraries and other organizations can use APIs to streamline services and maximize access.

Wade Guidry, the Library Technology Coordinator at the Collins Memorial Library at the University of Puget Sound, recently created several Yahoo Pipes applications that leverage data from the Times Best Sellers API and the OCLC WorldCat Search API.

Yahoo Pipes application

Each pipe retrieves a Times best-seller list, then pulls ISBN numbers from the API response and generates corresponding WorldCat queries. WorldCat reflects the holdings of libraries across the globe, making it easy for Collins Memorial Library patrons to zero in on the nearest available copy.

Best seller in WorldCat with holdings

Take a look at the pipes for hardcover nonfiction, paperback nonfiction and hardcover fiction. And then try out the resulting RSS links on the home page of the Collins Memorial Library catalog. In the RSS feeds, each best seller is a link to that title’s WorldCat record.

Feeds from Collins Memorial Library catalog page

With such innovative uses of data and APIs, the library has grown far beyond its four walls — so whether you live right next door to the Collins Memorial Library in Tacoma or hundreds of miles from any library at all, Times best sellers (and much more) are just a click away. We hope this example will encourage you to try your hand at creating some Yahoo Pipes with Times APIs. Be sure to let us know what you do with the data.

Our series of posts about external uses of Times APIs continues. Today we’re showcasing MovieMusicNow.com, which uses the Times Movie Reviews API.

Film may be primarily a visual medium, but the music of movies can be just as dazzling as 3-D (without the dorky glasses). Movie soundtracks can help you relive the drama, suspense or glee of a film — it’s like mood on demand. But it’s not always as "on demand" as you might hope: soundtracks can be difficult to find, especially for movies that aren’t marketed for their music. So what can you do if you’re hankering to hum along to the "lush dreamscape" of Avatar or the "flying body parts" of Ninja Assassin?

Movie Music Now can help you re-create the atmosphere of your favorite flick. Search for a movie, then sit back and listen to streaming samples, watch trailers and — our favorite part — read excerpts from Times movie reviews.

MovieMusicNow.com

The site uses APIs from Amazon, YouTube, The Times (specifically, the Movie Reviews API), Wikipedia and Twitter. Along with enjoying the streaming samples, you can check CD prices on Amazon and skim track lists. Each movie page also includes artist bios (from Wikipedia) and user reviews (via Twitter), and there’s even an option to post your own review.

User reviews and artist bios

Movie Music Now demonstrates the power and fun of combining several related sets of data into something greater than the sum of its parts. So spend some time perusing the site and the wide range of cinematic soundscapes on offer. If you’re not sure what you’re in the mood for, check the main page for new releases — the auditory assault of Alvin and the Chipmunks: The Squeakquel is just a click away. (Maybe some things really are better seen than heard.)

Our series of posts about external uses of Times APIs continues. Today we’re highlighting the Brooklyn Museum’s use of the Times Article Search API.

Remember those bumper stickers that said "Think Globally, Act Locally"? The slogan has gone out of style, but the sentiment is still worthwhile. It can even be applied to Times APIs: we hope they’ll be useful worldwide, but we’re thrilled when they pop up in our own backyard — say, at the Brooklyn Museum.

Last week, the Brooklyn Museum blog featured a post about recent efforts to digitize the museum’s wealth of resources, such as object labels and exhibition press releases. You might think that sounds like a lot of mind-numbing data entry and onerous OCR-ing — and it was, as the blog attests. But the museum’s intrepid tech staff also had a much less labor-intensive tool at its disposal: the Times Article Search API.

BrooklynMuseum.org and Article Search JSON

Developer Paul Beaudoin used the Article Search API to match Times articles to records in the museum’s exhibition archive. Find a show in the database, then check the "Press" tab for NYTimes.com links and excerpts. You’ll find everything from reviews and listings to editorials and glimpses of the social scene.

Times coverage of an exhibition

Press coverage is just one aspect of the rich layers of context the museum has added to its online offerings. For researchers, improvements like these go beyond "one-stop shopping"; the additional data can suggest other roads of inquiry or inspire related projects. And that’s exactly the kind of thing we were hoping our Article Search API could contribute to — because thinking globally often means leaving the thinking to others with great ideas.

For more on the digitization work (not to mention many other interesting topics), visit the museum’s blog. Be sure to check out the Collection API too, and follow @brooklynmuseum on Twitter. And if you’re in the neighborhood, visit the museum itself to see the real-world treasures atop the trove of data. Happy Holidays!

Brooklyn Museum Objects in the collection

Our series of posts about external uses of Times APIs continues. Today we’re highlighting Artist Explorer, which uses the Times Article Search API.

Earlier this year, we added a Microsoft Silverlight logo and link to the main page of the Times Developer Network. If you haven’t heard of Silverlight, it’s a framework that helps Web developers and designers create cross-platform applications.

Microsoft Evangelist Michael Scherotter created a Silverlight toolkit specifically for Times APIs. And while he was working on it, he took the opportunity to build a cool mashup of his own. Artist Explorer leverages Web services from The Times, Best Buy, Billboard and Intelepeer.

Michael S. Scherotter Artist Explorer

You can use Artist Explorer to research new artists or see what your old favorites have done lately. And when you find something you like, you can send SMS messages directly from the app — after all, your friends could probably use a little help with their holiday shopping lists.

The app uses the Times Article Search API, pulling in excerpts and thumbnails of articles related to the specified artist. Sort articles by date or headline, and click a facet category on the right to see Times-specific terms related to the artist: people, locations, organizations, subject terms (descriptions) and material types.

Michael S. Scherotter Article facets in Artist Explorer

On the Best Buy tab, links to Times reviews are displayed alongside product information, helping you make sure you really want to pay $32.99 for that Tom Waits box set.

Michael S. Scherotter NYT Reviews in Artist Explorer

Artist Explorer is an efficient, well-designed example of what you can do with Web services. (Others agree: the app won the Best in Show Billboard Award in the Circus Mashimus contest at the 2009 South by Southwest Interactive conference.) For more details, read Michael’s blog post or watch his screencast. We hope Artist Explorer will inspire you to experiment with the Silverlight toolkit for Times APIs — or will at least help you send out some holiday hints.

Our series of posts about external uses of Times APIs continues. Today we’re highlighting Infractor, a tangible user interface for the Times Article Search API.

A couple of years ago, I came across an article on Wired.com about Björk’s work with a multi-touch table called the Reactable. At first I wasn’t sure what to make of it: was this just another provocative prop, like that swan dress? But a little more reading revealed that the world of muti-touch tables and tangible user interfaces (TUIs) goes way beyond Björk — and is much more aesthetically appealing than that dress.

Multi-touch surfaces vary in form and function, but here’s the general idea of the Reactable: you place blocks or other objects on the surface and then manipulate them. The blocks interact with the surface and with each other, and you can also touch the surface with your fingers. Cameras and sensors track the objects (and fingers), and an application transforms the touches and movements into sound, light and data. (For more details, visit the Reactable site.)

The technology appeals to experimenters of all stripes, from musicians and artists to designers and developers. In the Interface Design department of the Potsdam University of Applied Sciences, a group of students recently found a way to use the Times Article Search API with a Reactable. They call their app "Infractor," but you could also call it "Mind-Blower."

Potsdam University of Applied Sciences The Infractor application: a tangible user interface for the Times Article Search API

Infractor allows you to search, filter and read Times article data. The app was built with two open-source tools: Processing and reacTIVision. Here’s a summary of how it works (from the Infractor site):

[A]ll available articles are placed as particles on the empty table. Each particle contains information such as the title, the actual text and further media like pictures and videos. This information is at that point not yet visible.

A source object is put in any place on the table in order to bundle the loose information. The object gathers all the particles and displays them in a ray. The particles move randomly inside of this ray. To be able to filter and screen them closer, the ray is split further with the help of prisms.

With controller objects, the prisms are given values that display thematically corresponding articles. Each controller object includes certain categorical values, such as individuals or countries. When such a controller slides close to a prism, a menu appears. By rotating the controller, it is possible to choose different values out of a category, [e.g.,] category: person; value: Obama. The chosen value is assigned to the ray. Another object, the magnet, allows one to see more information or the complete article.

This video gives you a better idea of the potential of Infractor, and makes it look both virtuosic and fun:

Infractor is a stunning example of what you can do with Times APIs. Kudos to Interactive Design students Stefanie Elsholz, Martin Schissler and Tillmann Dierichs, and to project supervisors Till Nagel and Professor Boris Müller. Here’s hoping Infractor inspires other dazzling experiments with Times data.

Our series of posts about external uses of Times APIs continues. Today we’re highlighting the Dallas Public Library’s use of the Times Best Sellers API.

Have you visited a library Web site lately? Maybe you think you have no reason to, especially if you’re not a regular patron of your local library. But on most library sites, you can do much more than look for a book: public branches offer everything from digital photo archives to podcasts to holiday cards. With so many services available, you might even forget to search the catalog.

Library services and Web services often go hand in hand. For example, the Dallas Public Library now uses the Times Best Sellers API to display best-seller lists.

The Dallas Public Library home page

This implementation illustrates the concept of making the available data work for you, in the way that best meets your needs. A typical <book> node in a Best Seller API response contains a lot of information:

<book>
   <rank>1</rank>
   <list_name>Trade Fiction Paperback</list_name>
   <bestsellers_date>2009-10-25</bestsellers_date>
   <published_date>2009-11-08</published_date>
   <weeks_on_list>9</weeks_on_list>
   <rank_last_week>0</rank_last_week>
   <asterisk>0</asterisk>
   <dagger>0</dagger>
   <book_details>
      <book_detail>
         <title>PUSH</title>
         <description>
            An abused, illiterate 16-year-old
            in Harlem meets a teacher who helps change
            her life; the basis for the film "Precious."
         </description>
         <contributor>by Sapphire</contributor>
         <author>Sapphire</author>
         <contributor_note />
         <price>12.95</price>
         <age_group />
         <publisher>Vintage</publisher>
      </book_detail>
   </book_details>
   <isbns>
      <isbn>
         <isbn13>9780307474841</isbn13>
         <isbn10>0307474844</isbn10>
         <isbn13>9780679766759</isbn13>
         <isbn10>0679766758</isbn10>
      </isbn>
   </isbns>
   <reviews>
      <review>
         <book_review_link>

http://www.nytimes.com/1996/06/14/books/

            books-of-the-times-a-cruel-world-
            endless-until-a-teacher-steps-
            in.html
         </book_review_link>
         <first_chapter_link />
         <sunday_review_link />
         <article_chapter_link />
      </review>
   </reviews>
</book>

Librarians Steve Gaither and Mark Gilman pared that down to the essentials, then added formatting, text and punctuation. The result is a clean, easy-to-skim list, complete with links to Times reviews and to the library’s own collection.

Best sellers with links to search the Dallas Public Library catalog

That’s what libraries are good at: making resources available to meet patrons’ needs, with minimum fuss and maximum usability. Stop by your own library’s Web site to see what it has to offer, and then stop by our Developer Network to see how you can make Times data work for you too.

Our series of posts about external uses of Times APIs continues. Today we’re highlighting Suburbified, which uses the Times Article Search API.

Here at NYTimes.com, the phrase "news and information platform" has become a common refrain. But sometimes catch phrases can lose their meaning, so let’s take a closer look at that one. The "news" aspect is what you might call a no-brainer. But what about "information" — isn’t that a little vague? Yes, intentionally so: quite a bit of the content you see on NYTimes.com isn’t strictly news and isn’t easily defined. From op-ed pieces to travel guides, our site offers a lot of material that goes beyond the headlines and covers a wide range of subjects.

And what can you do with that wealth of content? That’s where the "platform" part comes in: our APIs help developers build things with (and on) our data. They often seize the opportunity to zero in on something specific and present it in a new way. Suburbified.com illustrates that approach: it takes a particular subset of NYTimes.com "information" and transforms it into an annotated map to greener pastures.

Suburbified uses the Times Article Search API to gather locations and phrases from the "Living In" series of columns. Each week, these columns describe a New York City neighborhood or suburb. Using OpenStreetMap, Suburbified displays the profiled suburbs on a zoomable map.

Suburbified.com

Colors represent publication dates (redder = newer). Click the name of a town to see an excerpt and get the link to the full "Living In" article on NYTimes.com.

Summary and link

And Suburbified goes even further: it also provides a list of common "Living In" words and phrases. Scared of bears? Click that word on the right to see which towns to avoid on your road trips.

Common terms on Suburbified.com

Whether you live near rolling hills or strip malls, Times APIs can help you make the most of your online and offline neighborhoods. Stroll through Suburbified.com to see how news and information can take on new meaning — and maybe even lead to a change of address.

Our series of posts about external uses of Times APIs continues. Today we’re highlighting two sites that use the Times Newswire API.

The Times Newswire API doesn’t have a lot of bells and whistles. We decided to keep it simple and focus on one thing: helping you get up-to-the-minute content from NYTimes.com. Whether you call the API a fire hose, a stream or a shimmering waterfall, you know you’re getting a steady flow of updates — nothing more, nothing less.

But once you do have the stream of latest content, the sky’s the limit. You might decide to do something straightforward with the Newswire responses, such as posting links and summaries on your own news site. StateNews.com is an excellent example of this “pull in the headlines” approach.

StateNews.com

Published by students at Michigan State University, StateNews.com highlights campus goings-on, from Spartan Idol to professorial projects. But StateNews.com also offers a "Nation & World" section, powered by the Times Newswire API.

StateNews.com with Newswire data

The site takes the headline, byline and summary nodes of Newswire API responses and presents them in a two-column layout that’s easy to skim. Headlines are linked back to NYTimes.com. This may seem like an obvious use of the API, but that’s the whole point: it’s a clean, efficient way for the MSU community to get all the news of the university and the world, all in one place.

At the other end of the “imaginative uses” spectrum is Pretweeting.com, which uses the Newswire API for a game on Twitter. Built by Gabe Smedresman, the site is a virtual lexical stock market. Players buy and sell words with virtual money, based on how common they think the words will be on Twitter.

Pretweeting.com

The Newswire API is used in the research area of the site: click an individual word in an NYTimes.com headline or summary to see that word’s current value on Pretweeting. At the time of this writing, “health” was trading at 79 cents and “weekend” had soared to $2.10.

Pretweeting.com: researching words with Times Newswire

If you don’t quite grasp the point of Pretweeting.com, spend a few minutes on the site. You might get the itch to do a little day-trading — I’m already eyeing "costume," "candy" and "spooky." You can also use the research area of the site to see how (or whether) recent news stories are affecting the Twittersphere. And you can compare word trends at data.preweeting.com, which is a nice companion to our NYT Trender application.

From headline aggregation to virtual verbal empires, the Times Newswire API can feed every need for the latest NYTimes.com content. The bells and whistles are up to you.

Our series of posts about external uses of Times APIs continues. Today we’re highlighting two sites that use the Times Best Sellers API and the Times Movie Reviews API.

We’ve blogged about some complex uses of our APIs, including stunning visualizations and imaginative search applications. But sometimes simple is better: maybe you don’t want 12 ways to look at 12 different sets of data, and maybe you don’t even want to analyze the data at all. Maybe you’d just like to have fewer windows on your desktop and fewer mouse clicks in your day. In praise of simplicity, here are a couple of examples of straightforward ways to use Times APIs.

Audiobooks of Best Sellers, built by Hani Musallam, has a clear goal: to help you find audiobooks of New York Times Best Sellers. Using the Times Best Sellers API and the eBay APIs, it distills several sets of data into lists of available audiobooks. To start, select a list of best sellers; for each title on the list, you’ll see a cover image and brief description, along with the title’s current best-seller rank and the number of weeks it has been on the list.

AudiobooksOfBestSellers.com

Click a title to get a list of corresponding audiobooks on eBay and Half.com. Links to Times book reviews are also provided, so you can make sure you’re really ready to spend $4.99 on that audiobook of The Lost Symbol. You’ll also get details specific to eBay and Half.com, such as current bids, condition and shipping costs.

Audiobook listings

As Hani notes, AudiobooksOfBestSellers.com is still under development (for example, you’ll get better results from browsing the various best seller lists, rather than searching for a particular title). But the site already demonstrates that connections don’t have to be complicated to be useful. A single piece of data like an ISBN number can be the key to matching and combining data from several sources — and to helping you find what you need.

If you’d rather watch than listen, MovieAB.com is your one-stop resource. Built by Anurag Patel, it leverages the Netflix API and the Times Movie Reviews API to consolidate a wealth of movie information, including Times reviews.

MovieAB.com

For each movie, you’ll see basic data such as run time, release year and Netflix rating, as well as richer details like posters and pictures, cast and awards lists, and a link to the Times review (when available). Yes, you could get this information on other sites, but MovieAB.com pulls it all together for you. It’s especially useful for those times when you’re watching TV and thinking, “Where have I seen that guy before?” — and while you’re figuring that out, you can also see what Manohla Dargis or A.O. Scott thinks of that guy’s acting skills.

When you want to buy a CD of The Lost Symbol or skim the plot summary of The Da Vinci Code, these sites will help you with your to-do list. And you won’t even need a decoder ring.

About a year ago, we launched the Times Article Search API and the TimesTags API, two systems that allow developers to search our extensive collection of articles and to identify the canonical terms we use to describe them. Together, these APIs provide developers with the means to accurately find articles that are relevant to nearly any subject.

But there’s a problem with these (and most) APIs: they’re inherently narrow in their reach. The methods for using them are indecipherable for the average (non-developer) Times reader, and the data returned are formatted in a way that’s specific to each API, limiting their use across the Web. Enter the solution: custom RSS feeds.

Our new custom RSS tool is intended for all Times readers — not just developers. It provides a simple way to query the Times Article Search API and a standard way to consume the results.

Custom Times Feeds

You’ll find the tool at http://prototype.nytimes.com/customFeeds/. To use it, simply start typing something into the text field at the top. As you type, the application will try to match your words to our normalized list of terms. You can then either add one or more of the normalized terms to your feed (which will result in more targeted results) or add your original search phrase. (To add a term, click the arrow next to it to move it to the list of Your Feed’s Terms. To remove a term from your list, click the X next to it.)

Alternatively, you can paste the URL of an NYTimes.com article into the field at the top. You’ll then see a list of the terms our indexers used to describe that article, so that you can create a feed of similar articles.

Once you add a term to your feed, you’ll see a preview of the results, along with a message indicating the feed’s “strength.” The feed strength is determined by the number of articles published about your topic(s) within the past 30 days. Adding too many topics — or unrelated topics — can make a feed too specific; these “weaker” feeds produce fewer results.

Feed strength

You can continue to add terms and refine your feed until you’re happy with it. At that point, it’s simply a matter of naming the RSS feed and subscribing to it with your favorite reader. (If you don’t have a favorite RSS reader, you might try Google Reader, My Yahoo, Netvibes or My AOL.)

The options for creating a feed are intentionally limited — there’s no way to create a feed for one term OR another, for instance, only combinations of terms — in order to keep the application simple and approachable. Despite these functional limitations, however, the custom RSS tool excels at two very important tasks: following all of the news on NYTimes.com that’s relevant to a given topic, and following a particular news story as it develops over time. And it allows you to do these things in a familiar way — with RSS feeds.

The tool also incorporates the TimesTags API, which draws its power from a practice that is very rare nowadays: human meta-tagging. All New York Times articles are assigned keywords by real, live people, rather than by computers and search algorithms. And that means all the articles included in a custom feed are truly relevant to the topic specified. The TimesTags API also helps you follow a particular news story, because our producers and other information specialists apply consistent tags to all articles related to a given story.

We hope the custom RSS tool will demonstrate just how useful Times APIs can be — but more than that, we hope it provides a simple, customizable and familiar way to get news and information about almost anything you can think of. Take some time to play with the tool and create a few feeds. We think you’ll be pleased with the results.

Our series of posts about external uses of Times APIs continues. Today we’re highlighting the Elastic Times application, which uses the Article Search API.

The Times Article Search API defies description. On the one hand, the name of the API says it all: you use it to search for New York Times articles. That’s certainly its main purpose, and its parameters and responses are geared toward helping you find the articles you want.

But you can’t judge an API by its name, any more than you can judge a book by its cover. In addition to traditional search results, the Article Search API offers a rich set of data: facets. Think of facets as “perspectives” on search results — they allow you to approach your search from different angles. And because they’re a kind of metadata (or “data about the data”), facets are essential when you want to ask questions such as “When did the Sarah Palin coverage start to slow down?” or “Which European country is most often discussed in Times articles about oysters?”

Moritz Stefaner’s Elastic Times application helps you answer burning questions like those, and it also helps you formulate questions that may not have occurred to you when you first started searching. The app is an implementation of Moritz’s “elastic lists” prototype, which aims to minimize clicking and maximize results. (You might call that the holy grail for any search application.)

The Elastic Times application

Let’s try that oysters question as an example. A search for the keywords oysters and europe yields several useful facet sets, including descriptive subject terms (called “Description” in the app) and locations:

Facet sets

To see the “elastic” nature of the app, take a closer look at the facet lists. The individual segments stretch and shrink to convey proportion. In this case, we can quickly see that among articles about oysters, France is discussed more often than Germany: the value france takes up more space in the list than germany does.

Click a facet value to add it to your search query. You can even remove the keywords from your query and explore facets by themselves. Suppose you were to click great britain in the Location list and then delete your search terms from the keyword box. You’d start over with a general set of articles that are “about” Great Britain (as determined by skilled Times indexers) :

That might remind you of the trip you took to England a few years ago, so you might click travel and vacations in the Description list…

… and that might lead you right back to oysters.

As you use Elastic Times, the facet lists continue to stretch and shrink, guiding you to other perspectives and lending some elasticity to your brain as well as your search results. New facet values and results appear with each click, but the innovative design allows you to assess and reconfigure your query without corralling multiple windows or repeatedly clicking your browser’s Back button. The app harnesses two very powerful sets of data — facets and Times articles — but you won’t have to break a sweat to use it.

Considering the power of facets, we probably should have given the Article Search API a different name — something like Amazing Supreme Omnipotent Faceted Megasearch API. On second thought, apps like Elastic Times do a much better job of capturing the superlatives (and the stretchiness).