Predicting Activity and Location with Multi-task Context Aware Recurrent Neural Network

Abstract

Predicting users’ activity and location preferences is of great significance in location based services. Considering that users’ activity and location preferences interplay with each other, many scholars tried to figure out the relation between users’ activities and locations for improving prediction performance. However, most previous works enforce a rigid human-defined modeling strategy to capture these two factors, either activity purpose controlling location preference or spatial region determining activity preference. Unlike existing methods, we introduce spatial-activity topics as the latent factor capturing both users’ activity and location preferences. We propose Multi-task Context Aware Recurrent Neural Network to leverage the spatial activity topic for activity and location prediction. More specifically, a novel Context Aware Recurrent Unit is designed to integrate the sequential dependency and temporal regularity of spatial activity topics. Extensive experimental results based on real-world public datasets demonstrate that the proposed model significantly outperforms state-of-the-art approaches.

Predicting Human Mobility with Federated Learning

“…We remove POIs that have been visited less than 5 times and randomly select a subset of user access records and corresponding POIs information from each dataset. We apply Negative Log-Likelihood (NLL) [10], the prediction accuracy [3,16] and Average Percentage Rank [14,15] to evaluate the performance of our method. To demonstrate the effectiveness of STSAN, we compared with the following location prediction methods: ST-RNN [11], MCARNN [10], DeepMove [3], and VANext [4].…”

“…We apply Negative Log-Likelihood (NLL) [10], the prediction accuracy [3,16] and Average Percentage Rank [14,15] to evaluate the performance of our method. To demonstrate the effectiveness of STSAN, we compared with the following location prediction methods: ST-RNN [11], MCARNN [10], DeepMove [3], and VANext [4]. To demonstrate the effectiveness of federated learning model AMF, We compared with FedAvg [13] in our proposed STSAN method for location prediction task.…”

Location Prediction Models using Data Mining and Machine Learning

“…A Recurrent Neural Network (RNN) is a directed graph containing several artificial neural networks which for a network of several interconnected nodes exhibiting dynamic behavior. An RNN based model for activity and location prediction was suggested by Dongliang Liao, Weiqing Liu, Yuan Zhong, Jing Li, and Guowei Wang in [24]. The proposed method called MACARNN-Multi-task Context Aware Recurrent Neural Network used spatial activity for the prediction part.…”

A Fast Lightweight Spatiotemporal Activity Prediction Method

“…For example, the meaning of ''is there a restaurant'' is different from the meaning of ''there is a restaurant'', although the words contained in the two sentences are the same. Since the sequence model captures the text sequence features well, Liao D [18] proposed a novel Context Aware Recurrent Neural Network(RNN) to integrate the sequential dependency and spatiotemporal activity, Jung S [19] and Jiang J Y [20] used RNN to extract text sequence features and have achieved good results. Gated RNN [21]- [24] achieved state-of-the-art performance in sequential modeling.…”

“…Convergence: Defined by equation(18). We select s Accuracy thresholds, ep A i is the epoch used by method A to achieve Accuracy i , ep B i is the epoch used by method B to achieve Accuracy i .…”