experiments where the ranked retrieval lists from different runs are combined by rank-based re-scoring. The best retrieval per-.
確定！ 回上一頁
