This paper presents a method of estimating the level of difficulty for academic books using user reviews and the table of contents. Moreover, we build a simple user interface for recommender system based on difficulty level. As a choice of selecting academic books, the collaborative filtering has often been used for recommending them. However, the academic books selected by user preference are sometimes over user's knowledge resulting in wasting time and money. Hence, if we could know the difficulty level of the academic books which a user is browsing currently, the difficulty level available as meta data can be used for recommending them. Also, the difficulty level can be used as a parameter for selecting advanced books and for the user who gave up reading an academic book due to its difficulty.
