Programming Homework Help

Friend recommendation for a given user. Step 1: get data. Claw this user’s profile(introduction ,location ), this user’s recent tweeps(only text ,recent 3000 tweets, including contend, time, type). And same information for this user’s following user. And same information for this user’s following users’ following.(friend’s friend is a potential friend). For the data, it should be a csv file to store all information.(user_id, user_name,user_location, user_introduction, recent 3000 tweeps(tweep_content,only need text content, time to send this tweep, tweep type) ). Step 2,use keyword extraction for each user’s all tweeps to get 30 keyword for this user(tf-idf) . Calculate the friend’s friends’ 20 keywords to find which user’s keyword have high similarity with the target user. And calculate the distance of every user to target user’s location(if this user have location). Then do text similarity for users’ introduction to find which user have same interest with target user. Combined 3 elements(interest based on tweep content: keyword, distance between users, introduction between users) to get 5 most potential friends for the target user.