Name
Title | ||
|---|---|---|
A design space approach to analysis of information retrieval adaptive filtering systems |
Abstract | ||
|---|---|---|
In this paper we suggest a new approach to analysis and design of IR systems. We argue for design space exploration in constructing IR systems and in analyzing the effects of individual modules and parameters. We present results of experiments with parametric interpolation, or "homotopy", between two systems, and show, incidentally, that the best results are not achieved at the endpoints, and may lie outside the bounding hypercube defined by our choice of parameterization. Three distinct classes of interpolation are introduced to deal with the complexities of the specific example. |
Year | DOI | Venue |
|---|---|---|
2004 | 10.1145/1031171.1031228 | CIKM |
Keywords | Field | DocType |
distinct class,design space exploration,present result,individual module,best result,design space approach,specific example,new approach,information retrieval adaptive,parametric interpolation,ir system,adaptive filtering,homotopy,adaptive filter,information retrieval | Information retrieval,Parametrization,Computer science,Interpolation,Parametric statistics,Adaptive filter,Homotopy,Design space exploration,Hypercube,Bounding overwatch | Conference |
ISBN | Citations | PageRank |
1-58113-874-1 | 2 | 0.38 |
References | Authors | |
7 | 2 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Dmitriy Fradkin | 1 | 344 | 19.25 |
| Paul Kantor | 2 | 3 | 0.76 |