Robust Track

[摘要] In the robust track, we mainly tested our passagebased retrieval model withdifferent passage sizes and weighting schemes. In our approach, we used two retrievalmodels, namely the 2Poisson model using BM25 term weights and the vector spacemodel (VSM) using adaptive pivoted unique document length normalization. Also,we utilize WordNet to reweight some PRF terms and extract some context words asexpanded query terms. We show that our passagebased model achieves thecomparable performance on the whole query set. Moreover, our new methods ofusing WordNet information for query expansion can improve the retrieval

[学科分类] 社会科学、人文和艺术（综合）