Twitter turned on its long-awaited Geolocation API today, meaning that users can opt-in to having their messages annotated with their exact locations. The significance of this is made clear by comparing it with last week’s release of 500 million time-stamped Twitter messages for analysis.
// « You take this data, mash it up with any other very large corpus of data with timestamps, » Flip Kroner of data marketplace Infochimps told us, « and you’ve got a web app. » Today’s announcement of the availability of location data means the same thing: you take this data, mash it up with any other data with location information and you’ve got an app. From Digg or StumbleUpon for your favorite coffee shop to political and disease tracking – there’s a whole lot that’s possible.
Exposing location data is an opt-in feature for users, but 3rd party app developers are being told to « encourage your users to enable it by sending them to their settings page. »
Users will have to be both prompted and incentivized. Fortunately, a location-aware Twitter experience is something that will enable developers to deliver value to individual users immediately and in isolation – it doesn’t have to be one of those situations where « this will be cool once other people I know are using it. »
With the announcement today of Twitter search results being added to Yahoo News searches, Twitter data is now being used by all three of the major search engines. (Google’s implementation is still forthcoming, but the deal is done.) It might be one of the big players, but it’s more likely to be small innovators that make creative use of the new location data.
Twitter client Seesmic has already integrated geo data.These are possible Twitter use cases, but the standardized Activity Streams spec that Facebook, MySpace, Netflix and others now support also includes a geolocation field – so if the walls around Twitter ever fall to interoperability then we could be seeing innovations like these across all kinds of networks.
Here are some of the kinds of things we expect, or would like, to see.
« Party Over Here » Bot: Automated Geo-Replies
Want to know when you’re near a certain type of public event, great wine shops or deals at Macy’s? How about when friends, close friends or friends-of-friends are near? It’s not hard to imagine a bot that you subscribe to on Twitter, that then auto-subscribes to you, notices when you « check in » at a new location and automatically sends you a reply when whatever or whomever you’re interested in is near that location.
How about a bot you can Tweet « @whereami » to and that @’s you back with a link or stats about the location you’re in: nearby restaurant reviews, notable landmarks, crime rates, apartments for rent. Talk about augmented reality!There are all kinds of bots built on Twitter already, but one that can mash-up your physical location with its data store is going to be a lot more useful than a bot that tells you when a sensor noticed your plants need to be watered.
These are the kinds of services that will incentivize Twitter users to expose their location data. Assuming a substantial number of people make that choice, here are a few other examples that come to mind.
Articles Being Shared From This Coffee Shop Today Include…
Most Twitter search engines index not just the 140 characters in a message, but the text in links being shared as well. If you think people like being the Foursquare mayor of a popular coffee shop, imagine being the location-equivalent of Digg-submitter of your favorite coffee shop’s hottest online articles each day.Think people just stare at their computers in public these days? A service like this could shake that up. How about a StumbleUpon implementation that lets you stumble and read articles from people who’ve Tweeted from the same place you’re in. Imagine walking down the street and considering two competing coffee shops; what’s been on the reading list of each today?
News at 11: Local Interest Survey Tool
Think local TV news and newspaper companies would be interested in a stream of hot topics in their local area? They’d be foolish not to; what a great way to discover breaking local news to report on.
Does your local newspaper print a selection of letters mailed-in each week, but list the number of total letters received on the hottest topics? Imagine capturing that local chatter from a much larger sampling of people. Local tweets plus an entity extraction algorithm.
Imagine taking a map of tweets discussing criminal activity, or police misconduct, in a city and comparing it with a map of the same from local police agencies. Some places that warrant more official attention could be exposed.
If people in a certain city are twittering like fiends about a new product hitting the market, store orders, marketing and other parts of the supply chain could benefit from an earlier warning about it.
Politics & Marketing
People in Oregon are sharing a Huffington Post article about today’s health care reform announcement a lot? In Seattle, Washington perhaps not so much? Political organizers of a certain persuasion could find that information actionable.
Want to know what news outlets are on the ascent with people of a certain political persuasion? Cross reference your shared links from users in a location and a map of political contributions for the last election.How about unearthing Twitter users posting about environmental issues who also live in areas with environmental issues that an organization is working on.
Think Google’s use of search data to map out global disease trends is cool? Why stop there? How about pro-active messages (via Twitter) when there’s an increase in messages about being sick in your area?
Of course all of this will work better if more people are using Twitter and if people expose their location data, but that may very well happen. Prompting and individual incentives could be big drivers. The degree to which Twitter data is open for analysis by outside parties is a huge asset.
What would you like to see cross-referenced with Twitter location data?
Thanks for visiting ReadWriteWeb – we want to thank P2P-powered real-time search engine Faroo for making it possible for us to bring this site to you. Faroo is an innovative way to find out the hottest, freshest content on the web. Like SETI-at-home, Faroo’s distributed architecture is indexing the real-time web while ensuring user privacy by avoiding centralized storage of data. The company says it can do things with Chinese-language content that no other real-time search engine can, too. Check it out at Faroo.com.