The goal of the thesis is to implement a recommender system and a ASP.Net webshop, and to combine the two to visualize the results of the recommender system. The algorithm of the recommeder system is combining two algorithms and approaches from the two methods known from technical literature, Content-based filtering and Collaborative filtering, to get more precise results and to minimize the negative effects of the methods. The algorithm creates user and movie profiles, and then based on those it looks up the most similar movies and users on a given movie and user. Based on those subsets of the larger user and movie tables it creates a user-movie matrix, and then it makes a new matrix out of it with Singular Value Decomposition, where there will be the new, predicted rating of the user-movie pair. The recommender system has to work on any given database with the purpose of predicting ratings of movies that doesn’t break the given constraints. The database has to contain two tables, one of which has the movies and their attributes, while the other has ratings with the foreign keys of the movies and users. An ASP.Net webshop can use the database and show its results for the webshop’s users, if it has a database that has the two tables. The base of the ASP.Net webshop implementation is the base ASP.Net Visual Studio web project and doesn’t use any third-party webshop engine. On the webshop users can buy movies from a list and they can also rate them af the transaction. If the user hasn’t bought and rated a given amount of movies, the website predicts and recommends based on the average user scores of the movies. If the user has reached the target amount of ratings, the website with help of the recommender system makes predicted ratings and recommends based on those. The website is refreshing those predicted ratings if the user sends a given amount of new ratings.