id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
cord-353465-pej4e9z0	Ranjan, A.	FeverIQ - A Privacy-Preserving COVID-19 SymptomTracker with 3.6 Million Reports	2020-09-25		.txt	text/plain	3163	191	51	Unlike other trackers, FeverIQ uses secure multiparty computation (SMC) to cryptographically guarantee user privacy while providing insights to scientists and public health efforts. To address this need while cryptographically protecting the participant's privacy, the web application used SMC to determine four scores designed to capture the similarity of the user's symptoms to four preconfigured 'diagnosis' vectors: COVID-19-base, COVID-19-neuro, cold, and flu. We performed linear estimation on the inner product symptom scores for each of the four diagnosis vectors, by randomly dividing the data with complete input vectors and a reported test result into training and test sets (4:1 ratio). When we run the linear classifier on all participants who reported symptoms and provided complete input vectors, we found 3.4% were predicted positive, in line with results seen in seroprevalence surveys, such as 1.5% in Santa Clara County 24 , although publicly available positivity data vary widely.	./cache/cord-353465-pej4e9z0.txt	./txt/cord-353465-pej4e9z0.txt