Many enterprises attach more and more importance to online marketing. Among all kinds of the online advertising, keyword advertising catches most of the eyes. The research around keyword auctions becomes one of hottest topics. However, owing to the unique dynamic characteristics in keyword auctions, it's very difficult for the advertisers to find an effective strategy to keep the balance between the budget constraints and the effect of keyword bidding. Especially, it's very useful for them to find a global optimum bidding strategy.Based on the keyword auction products, this paper fully analyzes characteristic of the main keyword auction products and significance of research on global bidding strategy. It proposes a prediction method of keyword auction based on Bayesian network and establishes math functions for global optimum bidding strategies with different global and local constraints. By integrating the prediction method and functions, the research of the paper establishes a new kind of decision method and... |