Name
Abstract | ||
|---|---|---|
A large volume of user check-in data (check-ins) generated from location-based social networks enable a number of important location-aware services such as grouping users and recommending point-of-interests (POIs). Measuring the similarity between users according to check-ins is a key issue in many technologies for location-aware services such as clustering and collaborative filtering. Some works convert check-ins into vectors and compute the similarity between vectors, such as Cosine similarity and Pearson similarity, as the similarity between users. However, these similarity measurements do not exploit well the spatio-temporal gather and decay of check-ins. It can be easily observed that users tend to visit nearby places at nearby times. In this paper, we define co-occurrence patterns based on the time similarity and the location similarity. Then, we propose the spatio-temporal similarity by utilizing the most similar co-occurrence patterns. Finally, we verify the spatio-temporal similarity is effective by applying it to time-aware POI recommendation. |
Year | Venue | Field |
|---|---|---|
2018 | APWeb/WAIM Workshops | Data mining,Collaborative filtering,Social network,Cosine similarity,Computer science,Exploit,Cluster analysis,Temporal similarity |
DocType | Citations | PageRank |
Conference | 0 | 0.34 |
References | Authors | |
13 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Hongmei Chen | 1 | 25 | 5.39 |
| Peizhong Yang | 2 | 22 | 6.85 |
| Lizhen Wang | 3 | 153 | 26.16 |
| Qing Xiao | 4 | 7 | 3.52 |