Find similar books based on content. The python application uses postgresql and prolog custom prolog stemmer. We've downloaded the Chitanka's database and analysed the stemmed words with tf-idf algorithm. After that we determine the top 300 words for every book. Then we can recommend books as an answer to a query - the more common words they have in their top words with ours, the higher the score.