MAPS-KB

该数据集是一个覆盖了43万个三元组、涉及40万个术语的百万级概率性比喻知识库。它基于70GB的语料库构建而成，旨在评估比喻检测和成分提取方法的有效性。该数据集的规模达到了100万个比喻三元组，其任务是对比喻进行解释和生成。

This dataset is a million-scale probabilistic metaphorical knowledge base encompassing 430,000 triples and involving 400,000 terms. It is constructed based on a 70GB corpus, aiming to evaluate the effectiveness of metaphor detection and constituent extraction methods. With a scale of 1 million metaphorical triples, this dataset focuses on the tasks of metaphor interpretation and generation.