Office: INR 218
The spread of internet has made it possible for online feedback forums (or reputation mechanisms) to become an important source of information regarding the quality of products or services. Empirical studies show that buyers seriously consider online feedback when making purchasing decisions, and are willing to pay more for products with a good reputation.
Recent analysis, however, raises important questions regarding the ability of existing forums to reflect the real quality of a product. In the absence of clear incentives, users with a moderate outlook will not bother to voice their opinions, which leads to an unrepresentative sample of reviews. For example, Amazon ratings of books or CDs often follow bi-modal, U-shaped distributions where most of the ratings are either very good, or very bad. Under these circumstances, using the arithmetic mean to predict quality (as most forums actually do) gives the typical user a high variance estimator that is probably false.
The goal of this project is to find new methods for predicting the quality of a product from a given set of online reviews. For that we will use the numerical ratings given by different users, but also simple evidence from the textual comments that usually accompany a review.