search

feyclan / WT_analysis

For python cleaning and analysis
0 stars 0 forks source link

Get PoC for recommender systems working in python from the csv's #10

Open lucasschepers opened 6 months ago

lucasschepers commented 6 months ago

https://www.kaggle.com/code/yldzburhan/books-collaborative-filtering

https://www.kaggle.com/datasets/zygmunt/goodbooks-10k/data?select=tags.csv

https://www.kaggle.com/datasets/jealousleopard/goodreadsbooks/data

https://www.kaggle.com/code/hoshi7/goodreads-analysis-and-recommending-books

https://www.kaggle.com/code/sriharshavogeti/collaborative-recommender-system-on-goodreads

https://www.kaggle.com/code/omarzaghlol/goodreads-2-book-recommender-system

lucasschepers commented 6 months ago

Have some sort of demo where a book can be inputed and one (or several) models can recommend similar books based on user reviews. Techniques to be used may inculde collaborative filtering based on correlation coefficients, distances or cosine similarities. the best book may be found based on the (k-)nearest neighbours.

Because I don't know enough about unsupervised learning, I do not know how to evaluate the model with some sort of model performance. This needs be carefully considered before more advanced algorithms can be considered.