Optimization of Online Advertising Decision Combination Based on the Attribution of Consumer Behavior Big Data

Abstract

In the field of internet advertising, associated products based on various big data algorithms are widely presented on consumers’ browsing pages in order to increase their purchasing interest. With the help of computer big data processing, online advertisers can better grasp the shopping needs of consumers. This study proposes a chance constrained programming optimization model (CC-PR) for budget allocation of advertisers in a dynamic and complex consumer behavior context, based on the premise of analyzing consumers’ shopping journey from the perspective of the attribution of consumer behavior effects. It models different online behaviors generated by consumers’ interest transfer at different shopping stages. The characteristic of this method is to simulate changes in consumer psychology, overcome the difficulty of predicting multi-stage behavior of consumers, and extract a model of dynamic generalizable laws of consumer psychology. It dynamically adjusts advertising strategies based on changes in consumer tendencies at different stages. In order to verify the effectiveness of the proposed strategy, this paper introduces batch data from large shopping platforms and compares our method with the baseline method through experiments. The results show that the CC-PR method proposed in this paper has superior performance. A large amount of consumer behavior data can help advertisers make more accurate advertising decisions, and the method proposed in this article is also a good example of the application of computer big data processing in the field of online advertising.