NLP_Lyric_Clustering

Purpose of this project is to cluseter songs by some variables and try to come up with somekind of a relationship between them. There are 3 parts

Before even start,

1. You need to add a folder to project. Name it "data_set"
2. Download data from kaggle
3. Move data to "data_set" folder

Parts :

1. Info : Display info about data and preprocess a little
2. Feature Extraction : by using tokenizer, stemmer, deleting stop words, tfidf method and LDA, we will create vectors to feed to the clustering algorithm
3. Clustering : By using k-means, we will try to cluster data

Libraries needed to run:

1. NLTK
2. panda
3. numpy
4. sklearn

Todo:

1. MORE TESTS
2. PCA can be added
3. Current error rate is 14%. This can be improved.