Temporal collections of news articles (or news archives) contain numerous accurate and time-aligned articles, which offer immense value to our society, helping users to know details of events that occurred at specific time points in the past. Currently, the access to such collections is rather difficult for average users due to their large sizes and complexities. For better use of these valuable resources on our heritage, this study considers the task of machine reading at scale on long-term news article archives. We make use of the observation that questions on news archives are usually related to particular events and show strong temporal aspects. We propose a large scale question answering model designed specifically for long-term news article collections, with an additional module for re-ranking articles by using temporal information from different perspectives. The experimental results show that our model is superior to the existing question answering systems, thanks to dedicated module that allows finding more relevant documents.

新闻文章（或新闻档案）的时间收集包含大量准确且与时间保持一致的文章，这些文章为我们的社会提供了巨大的价值，可帮助用户了解过去特定时间点发生的事件的详细信息。目前，由于普通用户的规模大且复杂，因此很难访问此类馆藏。为了更好地利用这些宝贵资源来继承我们的遗产，本研究考虑了长期新闻文章档案中大规模机器阅读的任务。我们利用观察发现，新闻档案中的问题通常与特定事件有关，并且显示出强烈的时空方面。我们提出了一个专门针对长期新闻收藏而设计的大规模问答模型，带有一个附加模块，用于通过使用不同角度的时间信息对文章进行重新排名。实验结果表明，由于专用的模块可以查找更多相关的文档，因此我们的模型优于现有的问答系统。