Forest Reranking through Subtree Ranking

We propose the subtree ranking approach to parse forest reranking which is a generalization of current perceptron-based reranking methods. For the training of the reranker, we extract competing local subtrees, hence the training instances (candidate subtree sets) are very similar to those used during beam-search parsing. This leads to better parameter optimization. Another chief advantage of the framework is that arbitrary learning to rank methods can be applied. We evaluated our reranking approach on German and English phrase structure parsing tasks and compared it to various state-of-the-art reranking approaches such as the perceptron-based forest reranker. The subtree ranking approach with a Maximum Entropy model significantly outperformed the other approaches.