This post features a demo built using PredictionIO’s new Similar Product Template and dataset provided by Tapastic.com. Read on to find out more, you can also see how we built it with our step-by-step tutorial.
Tinder has quickly become one of the most successful and popular mobile applications largely due to its simple and addictive interface. Some may see it as simply a redux of “Hot or Not” but whatever your opinion no one can deny Tinder is very well designed and crafted user experience.
In fact Tinder’s success has spawned a number of spin offs such as Tinder for… news, jobs and even dogs!
Tapastic + Tinder = Tapster
Well we didn’t want to miss out on the fun! So we have open sourced a simple demo called Tapster a “Tinder for Comics” using data from Tapastic.com, the best place to read and discover comics online, and powered by PredictionIO, the open source machine learning server for developers and data scientists to build smart applications.
They Make You Laugh, They Make You Cry
Comics have the power both to delight and enrage. A “comic” quite literally means “causing or meant to cause laughter”. Whilst we can’t guarantee you will laugh we are going to cover how you can build a simple Tinder-like app using comics which have been provided by Tapastic.com.
Tapastic.com lets you enjoy visual stories and web comics, and even publish your own. They are all about “bite-sized, snackable content for your eyes to munch on”. Some of our favourite comics and comic artists include xkcd and Kal from The Economist’s but Tapastic.com has many more to discover.
Simply click the smiley green face to like and the red sad face to dislike a comic. Tapster will recommend new comics based on your interactions and which comics other users like and dislike.
How To Build Your Own “Tinder for X”
If you didn’t know already PredictionIO is an open source machine learning server for developers to build predictive features into their web and mobile app.
We provide Engine Templates for different predictive use cases to get you to production faster, in this case the Similar Products Template is powering the recommendation logic for which comic to show next based upon user’s likes and dislikes in real-time.
Th-th-th-that’s all folks… Have fun Machine Learning!
If you want to know how it was built then see our step-by-step tutorial https://docs.prediction.io/demo/tapster/