Cosmos QA：Machine reading comprehension with contextual commonsense reasoning
2019.09 -EMNLP 2019 論文 評測網站 介紹 Cosmos QA是一個35.6K問題的大規模數據集，需要基於常識的閱讀理解，被表述爲多項選擇題。 它着重於閱讀人們日常敘事的不同線條之間的界限，詢問有關事件的可能原因或影響的問題，這些事件需要推理超出上下文的確切範圍。 目前在此數據集上模型最高準確率爲 68.4%，相對於人類表現的94%。 閱讀理解的相關數據集 SQuAD 2016
相關文章
- 1. 【論文筆記】Knowledgeable Reader_ Enhancing Cloze-Style Reading Comprehension with External Commonsense Kg
- 2. Reasoning with Sarcasm by Reading In-between
- 3. r-net:machine reading comprehension with self-matching networks
- 4. 【論文筆記】Improving Machine Reading Comprehension with General Reading Strategies（2019，NAACL）
- 5. Unsupervised Deep Structured Semantic Models for Commonsense Reasoning
- 6. [Knowledge-based AI] {ud409} Lesson 15: 15 - Commonsense Reasoning
- 7. 論文淺嘗 - ICLR2020 | Abductive Commonsense Reasoning
- 8. 論文分享 - Reasoning with Memory Augmented Neural Networks for Language Comprehension
- 9. Attention-over-Attention Neural Networks for Reading Comprehension 訊飛
- 10. Simple and Effective Multi-Paragraph Reading Comprehension
- 更多相關文章...
- • XSLT
- • PDOStatement::fetch - PHP參考手冊
- • 爲了進字節跳動，我精選了29道Java經典算法題，帶詳細講解
- • 算法總結-股票買賣
相關標籤/搜索
每日一句
-
每一个你不满意的现在，都有一个你没有努力的曾经。
歡迎關注本站公眾號,獲取更多信息
相關文章
- 1. 【論文筆記】Knowledgeable Reader_ Enhancing Cloze-Style Reading Comprehension with External Commonsense Kg
- 2. Reasoning with Sarcasm by Reading In-between
- 3. r-net:machine reading comprehension with self-matching networks
- 4. 【論文筆記】Improving Machine Reading Comprehension with General Reading Strategies（2019，NAACL）
- 5. Unsupervised Deep Structured Semantic Models for Commonsense Reasoning
- 6. [Knowledge-based AI] {ud409} Lesson 15: 15 - Commonsense Reasoning
- 7. 論文淺嘗 - ICLR2020 | Abductive Commonsense Reasoning
- 8. 論文分享 - Reasoning with Memory Augmented Neural Networks for Language Comprehension
- 9. Attention-over-Attention Neural Networks for Reading Comprehension 訊飛
- 10. Simple and Effective Multi-Paragraph Reading Comprehension