Reading for next week

[Subbarao Kambhampati <rao@asu.edu>]

Catherine Wen is our guest lecturer for the next meeting. She will talk
about personalization systems.

She suggested the following as a reading for the talk:

Recommended paper:

@article{Scha-Kons2001,
        author = "J.B. Schafer
and J.A. Konstan and J. Riedl",
        year = 2001,
        journal = "Journal of
Data Mining and Knowledge Discovery",
        pages =
"115-152",
        title = "Electronic
Commerce Recommender Applications",
        volume = 5,
        number = 1/2,
        month = jan,
}

-       {Scha-Kons2001} http://www.cs.umn.edu/Research/GroupLens/ECRA.pdf

The term "recommender systems" evolved to replace and broaden the use of the term "collaborative filtering" because the latter term refers to a specific algorithm for recommending.  The term “recommender system” refers both to systems that specifically recommend lists of products and to those that help users evaluate products.  Previously, these systems were distinguished as systems that provide "recommendations," those that provide "predictions" of user preference, and those that provide "community opinion."  The recommender systems research community has embraced all three components.

In this paper, Schafer et al. surveyed the recommender applications used by several of the largest E-commerce companies. Several design parameters were identified and a taxonomy that classifies these applications by their inputs, output, recommendation method, degree of personalization, and delivery method was developed.  Classifying the applications revealed a set of application models that reflect the state of practice.  Promising directions in recommender systems have also explored, including application ideas built on innovative models that transcend current practice. 

Technologists often assume that the "best" recommender application is one that is fully automatic and completely invisible.  The study in this paper does not bear this assumption out at all.  Many different recommender models were found, each of which is appropriate for a different set of business goals.  Each of these models addresses different business needs that reflect different business models, customers, and marketing plans.  By selecting the appropriate combination of recommender applications, businesses can maintain their competitive advantage, retain customers, and increase sales.