Search

Obtaining fine-grained crowd flow distribution with recognized human intention is extremely valuable for a series of applications for a metropolitan city. We address this problem by leveraging a key insight people's intention behind their movement is highly correlated with the point-of-interest (POI) distribution of the corresponding regions. We propose DeepFlowGen to model the complicated relationship between crowd flow, POI, and time to generate intention-aware fine-grained crowd flow. Specifically, we solve the conflict between dynamic crowd flow and static POI distribution by fusing the information in both time and POI domains. In addition, we employ a sequence of residual blocks in DeepFlowGen to address the challenges of modeling the diverse temporal rhythms and heterogeneous influence of POI. Extensive experiments demonstrate that our model outperforms the state-of-the-art solutions by more than 15% in terms of RMSE of total crowd flow. Moreover, the correlation between the generated intention-aware crowd flow and the check-in distribution across different categories of POIs is as high as 0.90 and 0.75 in Beijing and Shanghai. Combined with extensive case studies, we demonstrate the strong ability of our model in generating intention-aware fine-grained crowd flow.

In this paper, we present a user-complaint prediction system for mobile access networks based on network monitoring data. By applying machine-learning models, the proposed system can relate user complaints to network performance indicators, alarm reports in a data-driven fashion, and predict the complaint events in a fine-grained spatial area within a specific time window. The proposed system harnesses several special designs to deal with the specialty in complaint prediction; complaint bursts are extracted using linear filtering and threshold detection to reduce the noisy fluctuation in raw complaint events. A fuzzy gridding method is also proposed to resolve the inaccuracy in verbally described complaint locations. Furthermore, we combine up-sampling with down-sampling to combat the severe skewness towards negative samples. The proposed system is evaluated using a real dataset collected from a major Chinese mobile operator, in which, events due to complaint bursts account approximately for only 0.3% of all recorded events. Results show that our system can detect 30% of complaint bursts 3 h ahead with more than 80% precision. This will achieve a corresponding proportion of quality of experience improvement if all predicted complaint events can be handled in advance through proper network maintenance.

Origin-Destination (OD) flow contains the information of direction and volume of population mobility between different regions in a city, having significant value in public transportation resource allocation. In this paper, we explore population distribution to infer OD flows, which is called pop2flow (population distribution to OD flows) problem. Compared to the conventional OD forecasting problem by using the historical OD matrix, pop2flow is more challenging because the population distribution carries much less information. In order to solve the pop2flow problem, we proposed a model, Graph-based Spatial-temporal Embedding with Dynamic Fusion (GSTE-DF). Specifically, GSTE-DF is composed of two parts: node embedding learning and flow prediction. The node embedding learning part captures the dynamic spatial-temporal features of population distribution into each nodes embedding. The flow prediction part adopts the learned embeddings and POI (points of interesting) distribution of every two regions to infer the population interaction between them. By conducting extensive experiments on real-world datasets collected in Beijing and New York City, we demonstrate the superiority of GSTE-DF compared to state-of-the-art baselines