The semantic similarity with original reports is high (Mean Weighted Overlap of 0.83). The precision and recall of extracted RTs based on experts' assessment ...
確定! 回上一頁