作者: Andrew S. Gordon , Christopher Wienberg , Sara Owsley Sood
DOI: 10.1109/SOCIALCOM-PASSAT.2012.43
关键词:
摘要: The utility of storytelling in the interaction between healthcare providers and patients is now firmly established, but potential use large-scale story collections for health-related inquiry has not yet been explored. In particular, enormous scale personal web logs offers investigators fields new opportunities to study behavior beliefs diverse patient populations outside clinical settings. this paper we address technical challenges identifying stories about specific health issues from corpora millions log posts. We describe a novel infrastructure collecting indexing posted each day English-language logs, coupled with user interfaces designed support targeted searches these collections. evaluate effectiveness search technology an effort identify hundreds first person third accounts strokes, purpose studying gender differences way that emergencies are described. Results indicate relevance feedback significantly improves search. conclude discussion sample biases inherent heightened by our approach, propose ways mitigate biases.