第3课--排序与CTR预估.pdfVIP

排序与CTR预估问题 七月在线 寒小阳 2017年1月7 日 Outline Ø  Online advertising and click through rate prediction Ø  Data set and features Ø  Spark MLlib and the Pipeline API Ø  MLlib pipeline for Click Through Rate Prediction Ø  Random Forest/GBDT/FM/FFM/DNN Online Advertising Online advertising Online advertising on search engines Social Media advertising Ø  Facebook's "Sponsored Stories" Ø  LinkedIn's "Sponsored Updates" Ø  Twitter's "Promoted Tweets" Online advertising is big business Types of online advertising Ø  Retargeting – Using cookies, track if a user left a webpage without making a purchase and retarget the user with ads from that site Ø  Behavioral targeting – Data related to user’s online activity is collected from multiple websites, thus creating a detailed picture of the user’s interests to deliver more targeted advertising Ø  Contextual advertising – Display ads related to the content of the webpage Online Advertising – The Players Publishers: Ø  Earn revenue by displaying ads on their sites Ø  Google, Wall Street Journal, Twitter Advertisers: Ø  Pay for their ads to be displayed on publisher sites Ø  Goal is to increase business via advertising Revenue models from online advertising Ø  CPM (Cost-Per-Mille): is an inventory based pricing model. Is when the price is based on 1,000 impressions. Ø  CPC (Cost-Per-Click): Is a performance-based metric. This means the Publisher only gets paid when (and if) a user clicks on an ad Ø  CPA (Cost Per Action): Best deal of all for Advertisers in terms of risk because they only pay for media when it results in a sale Click-through rate (CTR) Ø  Ratio of users who click on an ad to the number of total users who view the ad Ø  Used