Designing User-Centered Interactive Tools for Monitoring Software Learning Patterns

功能丰富的软件应用程序，如统计分析工具、图像和视频编辑器以及3D建模工具，可以包含数百个命令和功能，嵌入在精心设计的菜单和导航结构中。软件用户在与这些应用程序交互时表现出不同的学习模式。例如，一些用户试图遵循软件教程、视频或论坛中的特定说明，并在应用程序和网络之间来回切换数次。其他人不依赖外部资源，更喜欢通过试错来学习，以测试不同软件操作的效果。不幸的是，这些学习模式中的许多都是耗时且无效的，特别是对于无法轻松评估自己所做的是对还是错的新手用户。在拟议的研究中，我们将使用人机交互（HCI）的原理来发明新颖的工具和技术，允许用户在使用功能丰富的应用程序时自我监控，分析和反思他们的学习模式。自我监控的主要原则是，如果人们能更清楚地意识到自己在做什么，他们就能在行为上做出积极的改变。我们的长期目标是设计工具和算法，能够自动评估用户对新功能丰富的应用程序的学习情况，检测潜在的低效率，并提出有针对性的建议，以改善他们的交互。除了通过功能丰富的软件改善用户交互之外，我们的首要目标是使所有用户都能成为更高效、更自主的学习者。我们会把重点放在以下短期目标上，为实现长期目标奠定基础：(1)以用户为中心的交互式自我监控技术的设计和评估，允许用户在使用新的功能丰富的应用程序时可视化和反思他们的学习；(2)设计协作隐私敏感工具，让用户与个别专家或更大的用户社区分享他们的学习模式，以获得反馈和进一步反思；(3)聚合用户学习模式的技术，以帮助识别特定于应用程序的可学习性和可用性问题在进行这项研究时，我们面临的一个关键挑战将是考虑到软件使用和学习中的个体差异，并确保我们的工具和技术不仅快速准确，而且对不同的用户群体有用。此外，通过解决以有意义的方式聚合用户学习模式以获得大规模见解的挑战，我们的研究将为数据驱动的软件可学习性研究开辟新的途径。这项研究的好处将最终增强数百万软件用户的学习体验，并作为在广泛领域建立以用户为中心的学习干预的基础。

Feature-rich software applications, such as statistical analysis tools, image and video editors, and 3D modelling tools can contain hundreds of commands and features embedded in elaborate menu and navigation structures. Software users exhibit different learning patterns when interacting with such applications. For example, some users try to follow specific instructions in software tutorials, videos, or discussion forums and switch back-and-forth between the application and the web several times. Others do not rely on external resources and prefer learning by trial and error to test the effect of different software actions. Unfortunately, many of these learning patterns can be time-consuming and ineffective, especially for novice users who are not able to easily assess whether what they are doing is right or wrong. In the proposed research, we will use principles from human-computer interaction (HCI) to invent novel tools and techniques that allow users to self-monitor, analyze, and reflect on their learning patterns when using a feature-rich application. The main tenet of self-monitoring is that if people can become more aware of what they are doing, they will be able to make positive changes in their behaviours. Our long-term objective is to design tools and algorithms that can automatically assess how well a user is learning a new feature-rich application, detect potential inefficiencies, and make targeted recommendations for improving their interaction. Beyond improving user interaction with feature-rich software, our overarching goal is to empower all users to be more productive, self-directed learners. We will lay the groundwork for achieving our long-term objectives by focusing on the following short-term objectives: (1) user-centered design and evaluation of interactive self-monitoring techniques that allow users to visualize and reflect on their learning as they use a new feature-rich application (2) design of collaborative privacy-sensitive tools for users to share their learning patterns with individual experts or the larger user community for feedback and further reflection (3) techniques for aggregating users' learning patterns to help identify application-specific learnability and usability issues at scale In pursuing this research, a key challenge for us will be accounting for individual differences in software use and learning and ensuring that our tools and techniques are not only fast and accurate, but also useful across different user populations. Furthermore, by tackling the challenge of aggregating users' learning patterns in a meaningful way to derive large-scale insights, our research will open up new avenues for data-driven software learnability research. The benefit of this research will be in ultimately enhancing the learning experience of millions of software users and serving as a foundation for building user-centered learning interventions across a wide range of domains.