A reccomender system is developed using 1-page synopsys of thousands of books, titles and authors. A simple search engine using 'Levenshtein Distance' to search a book! A User-User and Book-Book co-similarity coefficients are calculated and User-Book matrix is generated. Using matrix factorization techniques such as PCA, ICA, LDA, user ratings for unseen books are predicted, and a top 10 recommendation are suggested.
The actual ipython notebook is in the Github repository (link below).