Obfuscation Remedies Harms Arising from Content Flagging of Photos

Obfuscation Remedies Harms Arising from Content Flagging of Photos

연구 분야: Analysis

학회: CHI '22: Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems

People share photos on Social Networks Sites, but at the same time want to keep some photo content private. This tension between sharing and privacy has led researchers to try to solve this problem, but without considering users’ needs. To fill this gap, we present a novel interface that expands privacy options beyond recipient-control (R). Our system can also flag sensitive content (C) and obfuscate (O) it (RCO). We then describe the results of a two-step experiment that compares RCO with two alternative interfaces - (R) which mimics existing SNS privacy options by providing recipient control, and a system that in addition to recipient control also flags sensitive content (RC). Results suggest RC performs worse than R regarding perceived privacy risks, willingness to share, and user experience. However, RCO, which provides obfuscation options, restores these metrics to the same levels as R. We conclude by providing insights on system implementation.