Hello, I have taken the time to drupalize movelens 1 million vote database. I am using it to test my recommendation engine. http://drupal.org/project/cre You can find the drupalized stuff and the java program that created it in my sandbox contrib/sandbox/scottreynolds I am posting here because, as one would expect, it takes awhile to query the database. Looking for ideas on how to speed it up. Here is the query that selects the recommended content. $db_result = db_query("SELECT d.content_id1 as 'content_id', sum(d.sum+d.count*r.value)/sum(d.count) as 'score' FROM {votingapi_vote} r, {cre_similarity_matrix} d WHERE r.uid=%d AND d.content_id1<>r.content_id AND d.content_id2=r.content_id AND d.content_type1='%s' AND d.content_type2=d.content_type1 AND r.tag='%s' GROUP BY d.content_id1 ORDER BY score DESC",$uid,$target_type,$tag); If you would like to get involved in the project or follow my progress or whatever ( i love to generate more interest) join my groups page: http://groups.drupal.org/soc-content-recommedation-engine