Twitter Pictures from Last Hour

One of the biggest issues currently facing social media platforms like twitter is how to best display the massive volumes of data that is generated with every new tweet.  The recent Iran protests were not covered by traditional media as they were banned from covering the situation, with images shared with services such as Twitpic.

TechCrunch also recently covered Twicsy calling it a Killer and Kind of Creepy Way to Search pictures, I was alerted to this service by an email from the founder Chris Seline.  Chris’s company Searchles is a social based search engine and its new product “Twicsy” allows for live searching of images posted using TwitPic or yFrog within the last hour.

The email I received outlining the Searchles spiel on the product and a few examples how you can use it to find recent images posted within the Twitter service.

“We just wanted to drop you a short not to let you guys know about Twicsy, a new search engine we just launched to search pictures that people post to Twitter. It is a great way to check out the most popular photos of the moment, and also breaking news in picture form. You guys have heard of the Twitter right? It’s all the rage.

(Caution, Twicsy is unfiltered, some images may be NSFW)

Some examples:
http://twicsy.com/?search=u2&sort=date (the u2 concert going on right now!)
http://twicsy.com/?search=dc+metro+crash
http://twicsy.com/?search=iran+election (caution! some graphic images)
http://twicsy.com/?search=rip+michael+jackson ”

The interesting part of human behaviour is the voyeur aspect combined with curiosity combined with almost OCD behaviour of not wanting not to miss out on something such as the recent Michael Jackson coverage has shown. While many other sites such as Facebook have encouraged with the ability to tag people in photos it does not allow you to find related images.  The Facebook tagging feature is great within social networks of friends and workmates, but how do you scale this to deal with breaking news stories such as Iran or Michael Jackson’s death and how do you expand and share these images easily?

Twicsy solves many of these issues and how they have built a time sensitive platform the benefit is they can react quickly to breaking news and less likely to have stale content as its refreshed continuously. The interface is easy enough when moving the mouse over images showing a larger version, more details such as tags and a link to what Twicsy discovered is likely related based on contextual tags.

This related images still seems to be in a ongoing learning process as shown in this example with a Twitpic of 2 girls in a bikini showing other similar photos, which many are not all be suitable for all audiences, but there are also several unrelated pics that are shown as apparently related:

• baked cookies
• coffee
• pet dog
• highway
• motorhome
• horse
• Amstel beer

The interesting social point is that it displays related tweets for that image and Twicsy allows you to quickly retweet the image.  This provides a great benefit as you can quickly share relevant images for breaking news stories, or viral campaigns can quickly spread within Twitter.

The interesting point about this product is that it doesn’t rely on the Twitter search API.  Twicsy indexes every tweet with an image link and then populates its service with this data.  Not using the Twitter search API means that it can develop and potentially improve its own image algorithm meaning its less likely to encounter spam as its search service improves.

The way that it indexes the image links outside the Twitter search API, Twicsy can potential build up a fresh image database to rival Google Images.  It also shows some extra features showing the product can potentially be scaled up shown by the nice little gem that is hidden in the right side menu is a popular pics of the past section with categories:
hour
2 hours
4 hours
12 hours
day
week

Twicsy also allows you to search its database for particular images, but many images are typically low resolution so not yet ideal to replace traditional media image archives it does provide a great way forward for better use of content found within social media. If you want to find out more about updates to their product follow Twicsy on Twitter.