We show that directional entailment is more helpful for inference than bidirectional similarity on questions of fine-grained semantics. We also ...
確定! 回上一頁