Home Technology Having trouble finding the perfect emoji to express yourself? Here’s the solution

0 1423

With thousands of expressive emojis and funny GIF reactions out there, it can be hard to choose the right one in the heat of the conversation. But now a new application Dango suggests which emoticon, gif, meme is best suited for what you are writing.  Dango, a new Android app from the Toronto-based startup Whirlscape, aims to make it easier to sift through the ever-growing array of choices after being launched for android on June 8. It uses innovative machine learning algorithms that are designed specifically to suggest users with appropriate emoticons.

The Dango app is named after a Japanese dessert consisting of sweet dumpling balls on a skewer (and yes, there’s an emoji for it). It sits atop other communication apps like Slack, Snapchat, or the built-in texting app on your phone. It suggests emoji that it thinks fit well with what you’re typing or in response to what someone just typed to you. You tap on the suggestion to add to the conversation.

Dango spins its emoji prediction game everywhere

According to its creators, Dango spins its emoji prediction game everywhere, no matter if you’re using Messenger, Kik, Whatsapp, Snapchat, or any other messaging app. It’s also compatible with any third-party keyboard apps you might be using, although it may still vary. If you face incompatibility, just shoot the developers an e-mail and help them support your keyboard.

Rather than simply coming up with word associations (such as a chicken emoji if you type “chicken”), Dango uses deep-learning techniques to try to figure out what whole sentences are expressing and then give you suggestions it thinks are related. For instance, if you type “She said yes!” Dango will show you the emoji for a ring and a bride with veil, among others.

In order to understand the meaning of what you are saying, Dango uses neural network- a computational structure with millions of adjustable parameters connected in ways similar to network of neurons in brain. The team created neural network by randomly initializing these parameters and then showing the network millions of real world examples of emojis use taken from across the web. At first the network just guesses randomly, but over time with each new training example, it slightly adjusts its millions of parameters so it performs better on that example.

While there are currently 1,624 emojis, this does not mean that there are only 1,624 meanings. When you use emoji, their meaning is determined by how they look and the context of their usage, which can be highly diverse. For example, a waving hand can mean hello, goodbye or simply a high five. This means the number of concepts Dango can represent is much greater than simply the number of individual emojis.

Currently, Dango is completely free and available only on Android. Further updates will probably introduce additional functionalities.


You can download Dango from here-


source: MIT technology review
Prerna Dhingra is a coder at heart who has just started her career in IT field. Eager to learn and explore new concepts are her area of interest. You can connect with her on Google+ and FB.


Leave a Reply