ECNU at 2015 CDS Track: Two Re-ranking Methods in Medical Information Retrieval

This paper summarizes our work on the TREC 2015 Clinical Decision Support Track. We present a customized learningto-rank algorithm and a query term position based re-ranking model to better satisfy the tasks. We design two learning-to-rank framework: the pointwise loss function based on random forest and the pairwise loss function based on SVM. The position based re-ranking model is composed of BM25 and a heuristic kernel function which integrates Gaussian, triangle, cosine and the circle kernel function. Furthermore, the Web-based query expansion method is utilized to improve the quality of the queries.