作者: Hiroshi Kanayama , Yusuke Miyao , Shuming Shi , Teruko Mitamura , Cheng-Wei Lee
DOI:
关键词:
摘要: This paper introduces an overview of the RITE (Recognizing Inference in TExt) task NTCIR-9. We evaluate systems that automatically recognize entailment, paraphrase, and contradiction between two texts written Japanese, Simplified Chinese, or Traditional Chinese. The consists four subtasks: Binary classification entailment (BC); Multi-class including paraphrase (MC); extrinsic application-oriented datasets: Entrance Exam RITE4QA. also describes how we built test collection, evaluation metrics, results submitted runs.