This is the repository for our research project: Evaluation of n-gram language models for lyrics retrieval
In this repository you can find the Jupyter Notebooks that were used to create the language models and retrieval functions. Preprocessing and model creation can be seen in Preprocessing.ipynb. The retrieval and re-ranking methods can be seen in Retrieval.ipynb.
The creation of language models takes a while because of all the data that must be parsed. In the following dropbox link you can download the required data for the Retrieval notebook. This folder contains pickle variable files which are loaded into the Retrieval notebook. Keep the files in the /Data directory relative to the notebook. (warning: 600MB)