Fig. 2: Probability of identification.
From: Interaction data are identifiable even across long periods of time
For each k ∈ {1, 2, 3}, we plot pk, the probability of identification within rank R ∈ {1, …, 43, 606} when the time delay is D = 1 week, with the 95% confidence interval shown in light blue. (Inset) shows the probability of identification for ranks 1, 10, and 100, with error bars for the 95% confidence interval. Our model correctly identifies people 52.4% of the time for k = 2. The probability of correct identification is still high at pk=1 = 14.7% for k = 1 and slightly increases pk=3 = 56.7% when k increases from 2 to 3. Our model ranks the correct candidate among the top 10 predictions pk=2 = 77.2% of the time and among the top 100 predictions pk=2 = 92.4% of the time for k = 2.
