Figure 6

(A) For each user we measure the fraction of places visited only once. This fraction is surprisingly large, as for each user on average 70% of the places were visited only once. (B) The probability of exploration estimated as fraction of explorations over the number of stops per user. This probability is quite large, meaning that on average users discover a new place every 4 or 5 stops. (C) Number of new visited (explored) places for each week, average by user. Surprisingly, the number of explored places does not decrease over time, but remains around 4. This highlights the highly exploring behavior of our population. (D) Cumulated number of new visited (explored) places for each week, average by user. As consequence of the large amount of exploration, the number of possible places to visit increases steadily over time, reaching on average almost 200 in one year. (E) Exploration prediction: \(f_{1}\) score for all models.