Going through a break-up can be difficult. Break-ups are emotionally fraught; two people have to untangle their lives. While prior research has paid attention to the implications of a break-up on the involved parties' mental and emotional well-being, and identified best practices for healing and growing from a break-up, these findings and practices are not well suited to how people's lives and experiences are represented in social media. People know how to break-up offline; however, they - and by extension, social media algorithms that recommend or remind - lack the ability to accurately represent the experience of breaking up online. Features that encourage connection or reminiscence are ill-suited for handling break-ups, making suggestions at a time when they may not be perceived as appropriate or desired. In my dissertation, I use three interpretivist, interview-based studies to investigate people's expectations of what is or is not appropriate behavior online after a break-up, and who to remain connected with on social media sites. My prior work established that algorithms are not always attuned to the nuances of human experiences like break-ups. Building on that work, my dissertation explores what people want when it comes to their online audiences and sharing behavior after going through a break-up, and leverages that understanding to offer design suggestions to improve both user experience and algorithmic systems in social media. My dissertation will result in empirical and design contributions that improve people's online experiences when going through a break-up.
