Lecture 12

This class was on the crowdsourcing of rankings via user comparisons.

Some relevant articles to read are:

• The Elo Ranking system used for chess on the Wikipedia.
• Various papers by Glickman (see attachments); the most important is the 1999 paper in which is described how to use normal-distribution approximations for producing the ranking.
• An interesting paper on ranking via ratings vs. ranking via comparisons.
• A very important paper on incentive structures for achieving a truthful ranking on the Web. The part that is important for us is Section 4.