Enhancing LambdaMART Using Oblivious Trees
This is an incremental improvement for search engine ranking systems.
The authors tackled the problem of improving LambdaMART for document retrieval by replacing standard regression trees with oblivious trees, resulting in a performance improvement of over 2.2%.
Learning to rank is a machine learning technique broadly used in many areas such as document retrieval, collaborative filtering or question answering. We present experimental results which suggest that the performance of the current state-of-the-art learning to rank algorithm LambdaMART, when used for document retrieval for search engines, can be improved if standard regression trees are replaced by oblivious trees. This paper provides a comparison of both variants and our results demonstrate that the use of oblivious trees can improve the performance by more than $2.2\%$. Additional experimental analysis of the influence of a number of features and of a size of the training set is also provided and confirms the desirability of properties of oblivious decision trees.