18936579. ADVERSARIAL BANDITS POLICY FOR CRAWLING HIGHLY DYNAMIC CONTENT (GOOGLE LLC)
ADVERSARIAL BANDITS POLICY FOR CRAWLING HIGHLY DYNAMIC CONTENT
Organization Name
Inventor(s)
Michael Bendersky of Cupertino CA (US)
Przemyslaw Gajda of Zurich (CH)
Marc Alexander Najork of Palo Alto CA (US)
Shuguang Han of Sunnyvale CA (US)
ADVERSARIAL BANDITS POLICY FOR CRAWLING HIGHLY DYNAMIC CONTENT
This abstract first appeared for US patent application 18936579 titled 'ADVERSARIAL BANDITS POLICY FOR CRAWLING HIGHLY DYNAMIC CONTENT
Original Abstract Submitted
Techniques of generating recrawl policies for commercial offer pages include generating a multiple strategy approach using a number of different strategies. In some implementations, each strategy is an arm of a K-armed adversarial bandits algorithm with reinforcement learning. Moreover, in some implementations, the multiple strategy approach also uses a machine learning algorithm to estimate parameters such as a click rate, impression rate, and likelihood of price change, i.e., change rate, which was assumed known in the conventional approaches.