使用基于注意的深度神经阅读器进行生物医学文本理解的初步研究:卡塔尔世界杯8强波胆分析设计与实验分析%A Kim, sesesoon %A Park, donghyun %A Choi,Yonghwa %A Lee,Kyubum %A Kim,Byounggun %A Jeon,Minji %A Kim,Jihye %A Tan,Aik Choon %A Kang,Jaewoo %+高丽大学信息学院计算机科学与工程系，首尔城北区安岩路145，韩国，82 02 3290 4840，kangj@korea.ac.kr %K机器理解%K生物医学文本理解%K深度学习%K机器理解数据集%D 2018 %7 05.01.2018 %9原始论文%J JMIR Med Inform %G英文%X背景:随着以深度学习为中心的人工智能(AI)技术的发展，计算机已经进化到可以阅读给定的文本，并根据文本的上下文回答问题的程度。这样一个特定的任务被称为机器理解任务。现有的机器理解任务大多使用一般文本的数据集，如新闻文章或小学水平的故事书。然而，没有人试图确定一个最新的基于深度学习的机器理解模型是否也可以处理包含专家级知识的科学文献，特别是在生物医学领域。目的:本研究旨在探讨机器理解模型是否能处理生物医学文章和一般文本。由于没有用于生物医学文献理解任务的数据集，我们的工作包括使用PubMed生成一个大规模的问题回答数据集并手动评估生成的数据集。方法:提出了一种适合生物医学领域的基于注意力的深度神经模型。为了进一步提高模型的性能，我们使用了预训练的词向量和生物医学实体类型嵌入。 We also developed an ensemble method of combining the results of several independent models to reduce the variance of the answers from the models. Results: The experimental results showed that our proposed deep neural network model outperformed the baseline model by more than 7% on the new dataset. We also evaluated human performance on the new dataset. The human evaluation result showed that our deep neural model outperformed humans in comprehension by 22% on average. Conclusions: In this work, we introduced a new task of machine comprehension in the biomedical domain using a deep neural model. Since there was no large-scale dataset for training deep neural models in the biomedical domain, we created the new cloze-style datasets Biomedical Knowledge Comprehension Title (BMKC_T) and Biomedical Knowledge Comprehension Last Sentence (BMKC_LS) (together referred to as BioMedical Knowledge Comprehension) using the PubMed corpus. The experimental results showed that the performance of our model is much higher than that of humans. We observed that our model performed consistently better regardless of the degree of difficulty of a text, whereas humans have difficulty when performing biomedical literature comprehension tasks that require expert level knowledge. %M 29305341 %R 10.2196/medinform.8751 %U http://medinform.www.mybigtv.com/2018/1/e2/ %U https://doi.org/10.2196/medinform.8751 %U http://www.ncbi.nlm.nih.gov/pubmed/29305341