By Henry Z. Tang. Stanford University.
Collaborative filtering (CF) is an ongoing development in the algorithms used for online recommendation systems that have become both a complement to and substitute for traditional search on online marketplaces. Most existing literature on the CF algorithm is understandably from an information systems standpoint, so this paper seeks to look at the economics behind this technological shift. It will examine namely how the distributions of online marketplaces have shifted and whether this shift favors niche products or large brand-name products. Are some marketplaces intrinsically more suited for collaborative filtering? What are the differences between an online mega-retailer such as Amazon and a subscription movie platform such as Netflix? After analyzing the Long Tail and Superstar effects in these marketplaces, the paper discusses broader implications for merchants, consumers, the platforms, and society as a whole.
Read the full paper here.