Automating the discovery of recommendation rules
D McSherry�- …�on Innovative Techniques and Applications of Artificial�…, 2004 - Springer
D McSherry
International Conference on Innovative Techniques and Applications of�…, 2004•SpringerWe present techniques for the discovery of recommendation rules that describe the
behaviour of a recommender system in localised areas of the product space. Potential uses
of the discovered rules include assessing the performance of the system in terms of
recommendation efficiency and solution quality. For example, the discovered rules may
reveal potential efficiency gains that might be achieved with an alternative recommendation
strategy. We also present an efficient algorithm for automating the discovery of�…
behaviour of a recommender system in localised areas of the product space. Potential uses
of the discovered rules include assessing the performance of the system in terms of
recommendation efficiency and solution quality. For example, the discovered rules may
reveal potential efficiency gains that might be achieved with an alternative recommendation
strategy. We also present an efficient algorithm for automating the discovery of�…
Abstract
We present techniques for the discovery of recommendation rules that describe the behaviour of a recommender system in localised areas of the product space. Potential uses of the discovered rules include assessing the performance of the system in terms of recommendation efficiency and solution quality. For example, the discovered rules may reveal potential efficiency gains that might be achieved with an alternative recommendation strategy. We also present an efficient algorithm for automating the discovery of recommendation rules in nearest-neighbour (NN) retrieval, the standard approach to product recommendation in case-based reasoning (CBR).
Springer
Showing the best result for this search. See all results