Millions of people are gathering long-term personal digital archives. We show how these archives may be used by consumers for their own benefit, by developing a system called Muse (Memories Using Email). Most directly, Muse helps people look back and reflect on the years or decades past, using their email archives. We have identified several types of cues that are effective in this task that are implemented in Muse. We also discovered that Muse is useful to support the work of archivists and researchers who often have access to the email archives of eminent individuals as part of libraries' special collections.

Apart from reminiscence, I'll introduce the idea of "experience-infused" applications, which are applications that can benefit by being connected to the user's digital archive. I will demonstrate two important examples embedded in Muse. The first is web search, where we create personalized search engines consisting of domains automatically collected from a user's social chatter. We show that these search engines can achieve surprisingly good performance, comparable to personalized Google search over the entire web. The second example is a digital archive-aware web browser that constantly scans terms on the current page, and inserts highlights and links to those that the user has encountered before. This browser is useful not only for personalizing busy web pages, but also in highlighting serendipitous connections. It makes realistic a powerful form of total recall.

A key feature of all these applications is that all data is analyzed on the user's behalf, and remains under his or her control, thus alleviating privacy concerns with service providers aggregating and monetizing the personal data of consumers.