Since emojis have been found to be a better tool to convey emotions rather than text, It is preferable to use them as a tool to retrieve related songs and music tracks.
What it does
Our Application receives a set of emojis from users and find the most related songs to them.
How we built it
A combinations of emojis descriptions and all corresponding html files are used as inputs for our python code. The code then uses cosine similarity algorithm to find a term with the highest probability of returning a song. For this purpose, the code is using a set of keywords with different weights to return the most accurate combination of emojis descriptions.
Challenges we ran into
First, Due to lack of a clean database containing emojis and corresponding descriptions, we had to clean a database to generate a new database to be used in our project. Second, producing a set of keywords has a critical importance for accurate suggestion.
Accomplishments that we're proud of
We are excited to have been able to successfully implement an application giving users music related emojis.
What we learned
We learnt how to use C# and python to create on the fly application.
What's next for Emotify
We would add a feature to give feedback to user to improve the results over time.