Itinerary recommendations dataset
- Citation Author(s):
-
Federica Cena (University of Turin)Luca Console (University of Turin)Marta Micheli (University of Turin)Fabiana Vernero (University of Turin)
- Submitted by:
- Fabiana Vernero
- Last updated:
- DOI:
- 10.21227/zk2x-pk70
- Data Format:
170 views
- Categories:
- Keywords:
Abstract
The dataset presents user evaluations for itinerary recommendations generated with three algorithms, PP, PP+TS and PP+TP.
Users evaluated recommendations according to five properties:
- overall evaluation (i.e., how good the recommended itinerary is overall),
- acceptance (i.e., how likely it is that the recommendation is actually accepted)
- organisation (i.e., how good the recommended itinerary is considering transfer times, number of PoIs, heterogeneity among PoIs, busy hours avoidance),
- PoIs (i.e., how good the recommended itinerary is as for the choice of PoIs to visit)
- free time (i.e., how good the recommended itinerary is as for the quantity of free time).
In the dataset, each row corresponds to a user. There are 124 users.
Instructions:
Each row corresponds to a user.
Columns correspond to properties:
UID_user - anonmous user ID
PP.overall - overall evaluation (i.e., how good the recommended itinerary is overall) for PP algorithm
PP.acceptance - acceptance (i.e., how likely it is that the recommendation is actually accepted) for PP algorithm
PP.organization- organisation (i.e., how good the recommended itinerary is considering transfer times, number of PoIs, heterogeneity among PoIs, busy hours avoidance) for PP algorithm
PP.POI - PoIs (i.e., how good the recommended itinerary is as for the choice of PoIs to visit) for PP algorithm
PP.free_time - free time (i.e., how good the recommended itinerary is as for the quantity of free time) for PP algorithm
PP+TS.overall - overall evaluation (i.e., how good the recommended itinerary is overall) for PP+TS algorithm
PP+TS.acceptance - acceptance (i.e., how likely it is that the recommendation is actually accepted) for PP+TS algorithm
PP+TS.organization - organisation (i.e., how good the recommended itinerary is considering transfer times, number of PoIs, heterogeneity among PoIs, busy hours avoidance) for PP+TS algorithm
PP+TS.POI - PoIs (i.e., how good the recommended itinerary is as for the choice of PoIs to visit) for PP+TS algorithm
PP+TS.free_time - free time (i.e., how good the recommended itinerary is as for the quantity of free time) for PP+TS algorithm
PP+TP.overall - overall evaluation (i.e., how good the recommended itinerary is overall) for PP+TP algorithm
PP+TP.acceptance - acceptance (i.e., how likely it is that the recommendation is actually accepted) for PP+TP algorithm
PP+TP.organization - organisation (i.e., how good the recommended itinerary is considering transfer times, number of PoIs, heterogeneity among PoIs, busy hours avoidance) for PP+TP algorithm
PP+TP.POI - PoIs (i.e., how good the recommended itinerary is as for the choice of PoIs to visit) for PP+TP algorithm
PP+TP.free_time - free time (i.e., how good the recommended itinerary is as for the quantity of free time) for PP+TP algorithm
Values range fro 1 (minimum) to 5 (maximum)
