Abstract | ||
---|---|---|
This paper proposes a novel method for assessing the performance of any Web recommendation function (ie user model), M, used in a Web recommender sytem, based on an off-line computation using labeled session data. Each labeled session consists of a sequence of Web pages followed by a page p$^{\rm ({\it IC})}$ that contains information the user claims is relevant. We then apply M to produce a corresponding suggested page p$^{\rm ({\it S})}$. In general, we say that M is good if p$^{\rm ({\it S})}$ has content “similar” to the associated p$^{\rm ({\it IC})}$, based on the the same session. This paper defines a number of functions for estimating this p$^{\rm ({\it S})}$ to p$^{\rm ({\it IC})}$ similarity that can be used to evaluate any new models off-line, and provides empirical data to demonstrate that evaluations based on these similarity functions match our intuitions. |
Year | DOI | Venue |
---|---|---|
2005 | 10.1007/11527886_44 | User Modeling |
Keywords | Field | DocType |
off-line computation,ie user model,corresponding suggested page p,associated p,web recommender sytem,off-line evaluation,page p,web recommendation function,session data,empirical data,web page,web pages,user model | Data mining,Combinatorics,Off line,Web page,Computer science,Intuition | Conference |
Volume | ISSN | ISBN |
3538 | 0302-9743 | 3-540-27885-0 |
Citations | PageRank | References |
1 | 0.37 | 7 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Tingshao Zhu | 1 | 192 | 33.61 |
R. Greiner | 2 | 2261 | 218.93 |
Gerald Häubl | 3 | 106 | 9.47 |
Kevin Jewell | 4 | 86 | 9.89 |
Bob Price | 5 | 481 | 31.72 |