Research on intrusiveness model of online advertising based on neural network

Online advertising has emerged as one of the major business models on the Internet. Publishers rely on the online revenue generated from online advertising to offer their free services. At the same time, intrusive ads such as pop-ups and animated layer ads are being perceived as annoying thing by an increasing number of Web users, because they distract the Web users from reading or navigating through the main content of web pages. It is definitely not a problem to place ads within Web pages as long as they do not become annoying. Thinking from the perspective of most users, online advertising should have a limit. It is necessary to model the intrusiveness of online advertising in order to ensure that a proper balance is achieved between user experiences and branding effectiveness. In this paper, an approach based on neural network is proposed to model the intrusiveness of online advertising. Comprehensive features of ads and hosting Web pages are exploited and effectively combined. Promising experimental results are obtained on Web pages collected from 500 Web sites.