Project information

• Category: Web, AI
• Project date: Spring 2018
• Github Repository: https://github.com/abradat/WordMap

Project Description
A project for generating word clouds for persian pop/rap songs and machine learning for predicting songs' genre.

Introduction

This project is implemented for Natural Languages Processing (NLP) course during my B.Sc. period. It has two parts:

1. For the first part, 107 pop songs and 59 rap songs’ lyrics were collected from Radio Javan. Normalization and Tokenizing the lyrics is done by hazm python module. after normalizing and extracting the words, we store the words and how many times they were repeated in the lyrics in the artist’s exclusive dictionary and category’s dictionary by founding out the exact number of words repeatition, we can now have our word maps.
2. For the second part, Deep Learning approach was used to detect the genre (Rap/Pop) of a given lyrics as the input to the neural network.

Technologies/Languages Used

Technology	Usage
	Python language is used for developing the applications
	Tensorflow is the framework for developing machine learning applications
	Git is used for the version control