Exploring temporal effects for location recommendation on location-based social networks
Citations
582 citations
Cites background or methods from "Exploring temporal effects for loca..."
..., Recall@k and Precision@k, in the top k POI recommendation [24, 12, 5]....
[...]
...Such a factorization algorithm has been exploited in [17, 5] for POI recommendation...
[...]
...For example, in [13, 25, 22], they tried to leverage content information of locations via topic modeling to assist POI recommendation; In [5], Gao et al....
[...]
520 citations
Cites background or methods from "Exploring temporal effects for loca..."
...Foursquare 3 [33] Check-ins, Friendships, User Three sets of check-ins from 33,596 users...
[...]
...Foursquare 3 [33] This dataset7 contains three datasets from Foursquare: a) check-in history of 18107 users, b) check-in history of 11326 users, and c) 4163 users who live...
[...]
...[33] further extends the model with using more aggregated temporal functions, such as sum, mean, maximum and voting, over the users’ check-in data....
[...]
340 citations
Cites background or methods from "Exploring temporal effects for loca..."
...• LRT: This is a recently developed matrix factorization method for POI recommendation with time information [7]....
[...]
...[7] study the temporal effect on the POI recommendation, but not the time-aware POI recommendation....
[...]
...By following previous studies [24, 7], we compute mtt∗ as:...
[...]
...Most of the existing POI recommendation methods [22, 3, 14, 9, 7, 12] overlook data scarcity and implicit feedback facts, and adapt conventional memory or model-based collaborative filtering for POI recommendation....
[...]
332 citations
Cites background from "Exploring temporal effects for loca..."
...As suggested in [7, 28], users’ mobility behaviors in the physical world exhibit strong temporal cyclic patterns, and the daily pattern (hours of the day) is one of the most fundamental patterns....
[...]
...By taking different definitions of temporal state, many other temporal patterns can be integrated into our GE model, as long as they contain the non-uniformness and consecutiveness properties [7]....
[...]
...[7] studied the temporal cyclic patterns of user check-ins in terms of temporal non-uniformness and temporal consecutiveness....
[...]
...Specifically, to overcome the data sparsity issue, most prior work of location-based recommendation focused on exploiting the geographical influence [10, 3, 20] and temporal cyclic effect [7, 28] to provide spatial or/and temporal context-aware recommendation....
[...]
References
11,500 citations
"Exploring temporal effects for loca..." refers methods in this paper
...Furthermore, the better performance of LRT than NMF suggests that time-dependent check-in preference capture user mobile behavior better than static check-in preferences....
[...]
...• Non-negative Matrix Factorization (NMF) Non-negative Matrix Factorization (NMF) [13] computes nonnegative user check-in preferences under the whole user-location matrix, which is our basic location recommendation model, as defined in Eq....
[...]
...Both NMF and R-LRT perform better than CF, demonstrating their ability in dealing with sparse data for location recommendation....
[...]
...Non-negative Matrix Factorization (NMF) Non-negative Matrix Factorization (NMF) [13] computes nonnegative user check-in preferences under the whole user-location matrix, which is our basic location recommendation model, as de.ned in Eq....
[...]
9,604 citations
9,583 citations
3,975 citations
"Exploring temporal effects for loca..." refers methods in this paper
...To solve largescale recommendation problems, matrix factorization is state-ofthe-art technology that has been proven to be successful in the Netflix Competition [11, 12], and is being used for item recommendation and trust prediction on product review sites like Epinions and Ciao for research purposes [15, 20, 21]....
[...]
2,922 citations
"Exploring temporal effects for loca..." refers background in this paper
...[5] proposed a Periodic & Social Mobility Model for location prediction with two temporal states (“home” and “work”) affected by social effects and non-social effects....
[...]
...Rm×n is a check-in indicator matrix, Yi j = 1 indicating correlations on LBSNs to solve the cold start location prediction that ui has checked in at lj, Yi j = 0otherwise. problem....
[...]
...Inspired by social influence theories that social friends tend to have similar check-in behavior, researches started to investigate the explicit social friendships on LBSNs [5, 9, 8] and leverage their power for improving location recommendation services [23, 3, 25]....
[...]
...However, these properties have not been exploited for location recommendation on LBSNs....
[...]
...Inspired by social in.uence theories that social friends tend to have similar check-in behavior, researches started to investigate the explicit social friendships on LBSNs [5, 9, 8] and leverage their power for improving location recommendation services [23, 3, 25]....
[...]