Enabling Author-Centric Ranking of Web Content

Abstract

In the past decade there has been a rapid increase in the amount of web content, partially due to wide-spread adoption of blogging and micro-blogging platforms that make it easy for ordinary users to quickly create web content. A single user may create content on many different online platforms or social networks. Current web search systems largely rely on link-structure of the web graph in ranking content. In other words, they care about “who links to a webpage” and are largely agnostic to “who authored the content on this webpage”. In this paper, we argue that in addition to link-structure the topic-authority of the original author can provide important signals for ranking and discovery of web content. We present algorithms for transferring topic-authority between webpages based on a same-origin policy. Our preliminary results suggest that such topicauthority transfers can improve the visibility of certain types of content compared to a link-structure only approach.