Book Crossing Data Cleansing for Recommendations
In the first notebook, "BookCrossing data cleansing", I will use some techniques to perform Data Exploration and Cleansing on the Book-Crossing Dataset collected by Cai-Nicolas Ziegler. In doing so, I aim to gain some intuition about the data. I also aim to prepare the data so as to be used in a recommender system and provide book recommendations. The most used weapon of choice for this notebook will be Pandas library.
The second notebook named Recommender uses the recommender_class.py to create a command-line user interface. The user is able to rate items and get recommendations based on them. Since the dataset is sparse it works best if the user rates a lot of items. To provide the predictions we use surprise library (http://surpriselib.com/).
I hope you enjoy the project and the whole experience. If you'd like to contact me, my email is "ppelitaris@gmail.com".