MS-LaTTE

MS-LaTTE数据集是由微软研究院雷德蒙德分部创建的一个大规模真实生活任务完成数据集，包含10,101个任务，是迄今为止最大的公开可用任务数据集。该数据集特别之处在于，它不仅记录了任务本身，还包含了任务完成时的位置和时间标签，这为研究任务完成的上下文提供了重要信息。数据集的创建过程涉及从Wunderlist应用的日志中抽取任务，并通过内部众包平台进行位置和时间的标注。该数据集的应用领域广泛，主要用于研究任务智能，特别是在预测任务的空间和时间共现性方面，有助于开发更智能的任务管理工具。

The MS-LaTTE dataset is a large-scale real-life task completion dataset developed by the Redmond division of Microsoft Research, consisting of 10,101 tasks and standing as the largest publicly available task dataset to date. What distinguishes this dataset is that it not only records the tasks themselves but also includes location and time stamps at the time of task completion, which provides critical information for researching the context of task completion. The dataset was created by extracting tasks from the logs of the Wunderlist application and annotating their locations and timestamps via an internal crowdsourcing platform. It has a wide range of application scenarios, primarily used for research on task intelligence, especially in predicting the spatial and temporal co-occurrence of tasks, which facilitates the development of more intelligent task management tools.