| 冯娟
"PhD candidate in Management Science & Information Systems with Operations Research Dual Title Penn State University "
|
|
【主题】Paid Placement in Information Gatekeepers
【时间】2003-3-20上午10:00 【地点】清华经管学院 北207
【语言】中文/英文
【内容摘要】
Information gatekeepers-such as Internet search engines, shopbots, financial advisors, referralites, and online travel services-are an essential entry point for many information earch and decision making tasks. An information gatekeeper uses its repository of relevant information and algorithms to provide users a ranked recommendation list in response to a specific query. This paper analyzes the practice of paid placement in information gatekeepers, where the gatekeeper biases its outputs to favor certain providers who pay it a placement fee. We examine a gatekeeper's bias decisions, how these are affected by the elasticity of demand towards bias, and the interplay between bias and technology aspects of gatekeeper's quality. A monopoly gatekeeper will choose independence only when the users' sensitivity to bias is very high. An increase in technological quality will cause the gatekeeper to lower its bias level when the lasticity of marginal demand for quality is lower than the per-user contribution of placement revenue. Competition will cause gatekeepers to reduce bias level. For heterogeneous gatekeepers, the one with higher technology level may or may not choose a lower bias level than opponents, but will capture a larger market share. In choosing the technology level, the gatekeeper with superior cost function will set a higher technology level. Its optimal bias level will be such that its composite quality measure is better than its opponents. The competitive choices of bias level may in some cases be Pareto inefficient. Simulations about different ranking mechanisms used by major search engines such as Google and Overture are conducted and a better mechanism is proposed. |
