This paper focuses on the adoption of Agent Technology to calculate and evaluate the Web Content Engagement Time (WCET). Traditional Web traffic analysis metrics such as pageviews, unique browser, visitor loyalty, etc have been used to analyze the web traffic behaviour for a long time since the birth of World Wide Web, but the emersion of software robots and web crawlers trigger a huge impact on the integrity and correctness of these traditional Web statistics. For advertisers, these statistics are not enough for them to evaluate the actual return-on-investment (ROI). For instance, large amount of pageviews but extremely short session duration will not have much impact for the advertisers to promote their products and brands. Web Content Engagement Time (WCET) for the reading on interactive Web content such as e-magazines and e-publications, which focuses on the page duration between each "page-flipping", will give advertisers much more information and confidence on whether such eye-balls (i.e. attention) are actually focused on the Web content (and hence the eAds) or not, especially during the browsing of e-magazines and e-publications. But such indicator involves significant amount of calculation within the Web server, especially when over thousands of users are reading a popular e-publication at the same time. To tackle with this problem, a multi-agent based Web Content Engagement Time (WCET) Analyzer is proposed on e-publication system. From the experimental perspective, popular Chinese e-magazine "MingPaoWeekly", with over 0.5 million readership in Hong Kong and overseas Chinese communities are tested over the IAToLife.com Web Channel platform, promising Web Content Engagement Time (WCET) is recorded, which provides not only integrity and confidence for the publishers and advertisers, but also shines a new light for the future agent-based target marketing and e-reader profile and reading behavior analysis.
展开▼
