A major criticism of peer-to-peer (P2P) networks,
such as Napster and Gnutella, is that their main
purpose is the pirating of intellectual property.
However, a redeeming aspect of P2P networks is
that they can perform a valuable promotional function
by making users aware of a product that they
would not otherwise have considered. Specifically,
peers can recommend a song to another peer. Music
tends to have a great deal of heterogeneity in
tastes, and if peers are connected to peers with
similar preferences then the recommendations
serve a valuable promotion function. In this research
we seek to understand how P2P networks
could aid information producers in increasing the
user base of their products and their profitability.
Traditionally producers have relied upon broadcasting
playlists that are targeted towards a homogeneous
set of users, which can limit their
value. A potential benefit of P2P systems is that
they can serve a collaborative filtering function by
correlating preferences across users and providing
playlists customized to an individual user’s tastes.