Human-Computer Interaction Logs

This dataset comprises ten human-computer interaction logs of real participants who solved a given task in a Windows environment. The participants were allowed to use the standard notepad, calculator, and file explorer. All recordings are anonymized and do not contain any private information. Simple:Each of the five log files in the folder simple contains Human-Computer Interaction recordings of a participant solving a simple task. Participants were provided 30 raw text files where each one contained data about the revenue and expenses of a single product for a given time period. In total 15 summaries were asked to be created by summarizing the data of two files and calculating the combined revenue, expenses, and profit. Complex:Each of the five log files in the folder complex contains Human-Computer Interaction recordings of a participant solving a more advanced task. In particular, participants were given a folder of text documents and were asked to create summary documents that contain the total revenue and expenses of the quarter, profit, and, where applicable, profit improvement compared to the previous quarter and the same quarter of the previous year. Each quarter’s data comprised multiple text files. The logging application that has been used is the one described inJulian Theis and Houshang Darabi. 2019. Behavioral Petri Net Mining and Automated Analysis for Human-Computer Interaction Recommendations in Multi-Application Environments. Proc. ACM Hum.-Comput. Interact. 3, EICS, Article 13 (June 2019), 16 pages. DOI: https://doi.org/10.1145/3331155 Please refer to Table 1 and Table 2 of this publication regarding the structure of the log files. The first column corresponds to the timestamp in milliseconds, the second column represents the event key, and the third column contains additional event-specific information.

本数据集包含10份真实参与者在Windows环境下完成指定任务的人机交互（Human-Computer Interaction）日志。参与者可使用标准记事本、计算器与文件资源管理器。所有录制内容均已完成匿名化处理，未包含任何私人敏感信息。

简单任务子集：文件夹simple内的5份日志文件，每份对应一名参与者完成简单任务的人机交互录制内容。参与者将获得30个原始文本文件，每个文件包含某一特定时间段内单一产品的收支数据。参与者需通过汇总两份文件的数据，计算合并后的总收入、总支出与总利润，最终生成总计15份总结文档。

复杂任务子集：文件夹complex内的5份日志文件，每份对应一名参与者完成更高阶任务的人机交互录制内容。具体而言，参与者将收到一个包含若干文本文档的文件夹，需生成包含以下内容的总结文档：季度总收支、利润，以及（适用场景下）相较于上一季度与去年同期的利润增幅。每个季度的数据由多个文本文件构成。

本数据集采用的日志录制工具来自如下文献：Julian Theis与Houshang Darabi. 2019. 多应用环境下基于行为Petri网挖掘与自动分析的人机交互推荐方法[C]//Proceedings of the ACM on Human-Computer Interaction. 3, EICS, Article 13 (June 2019), 16 pages. DOI: https://doi.org/10.1145/3331155。

请参考该文献中的表1与表2，以了解日志文件的结构。日志文件的第一列为以毫秒为单位的时间戳，第二列为事件键值，第三列包含事件专属的附加信息。