Teaching Machine Comprehension with Compositional Explanations
Department of Computer Science, University of Southern California（南加州大学计算机科学系）；Information Science Institute, University of Southern California（南加州大学信息科学研究所）
Qinyuan Ye, Xiao Huang, Elizabeth Boschee, Xiang Ren
Advances in machine reading comprehension (MRC) rely heavily on the collection of large scale human-annotated examples in the form of (question, paragraph, answer) triples. In contrast, humans are typically able to generalize with only a few examples, relying on deeper underlying world knowledge, linguistic sophistication, and/or simply superior deductive powers. In this paper, we focus on "teaching" machines reading comprehension, using a small number of semi-structured explanations that explicitly inform machines why answer spans are correct. We extract structured variables and rules from explanations and compose neural module teachers that annotate instances for training downstream MRC models. We use learnable neural modules and soft logic to handle linguistic variation and overcome sparse coverage; the modules are jointly optimized with the MRC model to improve final performance. On the SQuAD dataset, our proposed method achieves 70.14% F1 score with supervision from 26 explanations, comparable to plain supervised learning using 1,100 labeled instances, yielding a 12x speed up.
TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension
Paul G. Allen School of Computer Science & Engineering, Univ. of Washington（华盛顿大学保罗·G·艾伦计算机科学与工程学院）；Allen Institute for Artificial Intelligence（艾伦人工智能研究所）
我们提出了TriviaQA，一个具有挑战性的阅读理解数据集，包含超过650 K的问题-答案-证据三元组。TriviaQA包括由琐事爱好者撰写的95 K问答对和独立收集的证据文件，平均每个问题六个，为回答问题提供高质量的远程监督。我们发现，与最近引入的其他大规模数据集相比，TriviaQA（1）具有相对复杂的组成问题，（2）在问题和相应的答案证据句子之间具有相当大的句法和词汇变异性，（3）需要更多的跨句子推理来找到答案。我们还提出了两个基线算法：基于特征的分类器和最先进的神经网络，表现良好的SQuAD阅读理解。这两种方法都没有接近人类的表现（23%和40% vs. 80%），这表明TriviaQA是一个具有挑战性的测试平台，值得在未来进行大量研究。数据和代码可在--http://nlp.cs.washington.edu/triviaqa/获得
Testing AI on language comprehension tasks reveals insensitivity to underlying meaning
Universitat Rovira i Virgili（罗维拉-威尔吉利大学）；Humboldt-Universitat zu Berlin（柏林洪堡大学）；UTHealth（德克萨斯大学健康科学中心）；New York University（纽约大学）；Universitat Autònoma de Barcelona（巴塞罗那自治大学）；Institució Catalana de Recerca i Estudis Avançats（加泰罗尼亚高等研究院）
One Size doesn't Fit All: A Personalized Conversational Tutoring Agent for Mathematics Instruction
School of Computer Science, Wuhan University（武汉大学计算机学院）；DAMO Academy, Alibaba Group（阿里巴巴达摩院）；College of Computer and Cyber Security, Hebei Normal University（河北师范大学计算机与网络空间安全学院）
针对数学教学中学习者差异被忽视的问题，本文提出 PACE 个性化对话辅导智能体，结合学习风格画像、苏格拉底式反馈与个性化训练，提升教学适应性和学习效果。
