Model Adaptation via Model Interpolation and Boosting for Web Search Ranking

TL;DR

This paper compares model interpolation and boosting methods for web search ranking adaptation, finding that interpolation excels on diverse data, while boosting performs well on similar data but struggles with data variability.

Contribution

It introduces and evaluates model interpolation and boosting approaches for web search ranking adaptation, highlighting their strengths and limitations in different data scenarios.

Findings

Model interpolation outperforms on diverse, open test sets.

Boosting performs best on similar, closed test sets.

Boosting's performance drops on open test sets due to instability.

Abstract

This paper explores two classes of model adaptation methods for Web search ranking: Model Interpolation and error-driven learning approaches based on a boosting algorithm. The results show that model interpolation, though simple, achieves the best results on all the open test sets where the test data is very different from the training data. The tree-based boosting algorithm achieves the best performance on most of the closed test sets where the test data and the training data are similar, but its performance drops significantly on the open test sets due to the instability of trees. Several methods are explored to improve the robustness of the algorithm, with limited success.