Streamlined Tools for Personal Analytics

点击翻译按钮获取中文摘要

This Discovery Grant will focus on the design of new tools and applications that enable "logging at the speed of thought"—allowing people to capture detailed, frequent observations about their thoughts and activities without consuming significant time or attention. The recent wave of personal sensors and self-tracking tools has promised to help people use data to better understand themselves. By making it possible to track our physiology and our activities, self-tracking tools should allow everyday users to measure, understand, and deal with issues like chronic pain and emotional health, or support ongoing goals like learning a new language. However, easily automated measurements like step counts and application logs are often too low-level to answer these kinds of questions. Meanwhile, current mobile applications for manual logging are often so cumbersome, time consuming, and incompatible with daily activities that people fail to keep consistent or accurate records. However, many important pieces of personal data—including pain logs, emotion journals, and notes on language use—require human judgement and observation and must be recorded manually. My research will explore tools and techniques for streamlining personal data collection and analysis. By minimizing the overhead associated with self-logging, our prototypes will make it possible to collect and analyze much greater quantities of rich personal data. These tools will allow people to quickly record information in the context of their daily activities and thoughts (Where am I experiencing pain and how intense is it? Did something change in my environment? Did I just mispronounce that word?) at times when it would otherwise be difficult to do so. Lightweight recording tools will help people record activities without losing focus on their current task—making it possible to record and analyse more types of activities, and also supporting applications like pain tracking where logging ideally shouldn't focus the person's attention on symptoms as they are occurring. My team’s research will examine these challenges through the design of a toolkit for streamlined data collection and analysis. We will design minimalist logging approaches based on speech and touch that are as smooth and reliable as possible, and explore subtle feedback mechanisms that help people detect errors and add structure to support later analysis. These logging tools will be backed by tailored processing and analysis tools that allow people to easily filter, clean, and organize their logs before analysis and build reusable analysis pipelines that they can use to examine future data. Finally, we will explore the potential of these lightweight logging tools in a series of domain-specific applications, including tools for pain management and language learning.