Who Am I?

To enable targeted ads, companies profile Internet users, automatically inferring potential interests and demographics. While current profiling centers on users' web browsing data, smartphones other devices with rich sensing capabilities portend techniques that draw methods from ubiquitous computing. Unfortunately, even existing ad-targeting practices remain opaque to engendering distrust, resignation, privacy concerns. We hypothesized making visible at the time place it occurs might help users better understand engage constructed profiles. this end, we built a technology probe surfaces incremental construction of user profiles both activities in physical world. The explores transparency control real time. conducted two-week field deployment 25 participants. found increasing visibility helped participants anticipate how certain actions can trigger specific ads. Participants' desired engagement their differed part based overall attitudes toward Furthermore, expected algorithms would determine when an inference was inaccurate, no longer relevant, or off-limits. Current typically do not this. Overall, our findings suggest leveraging opportunistic moments within pervasive computing own inferred create more trustworthy positive experiences