Xinyan Zhu

Repeat and near-repeat burglaries and offender involvement in a large Chinese city

Repeat and near-repeat phenomena indicate that victimization might form a contagion-like pattern in both near space and near time. The research community has built abundant empirical evidence of repeat and near-repeat patterns over the past 10 years. Along with these pattern analyses, scholars have also proposed theoretical explanation of the mechanisms of near-repeat burglaries. However, the empirical study of mechanisms is very limited and almost absent in China due to the data unavailability. Utilizing a recent burglary dataset in Wuhan (a large Chinese city), the current research examines repeat and near-repeat burglary patterns as well as the underlying theoretical hypotheses. Consistent with prior findings, the current study found significant and meaningful near-repeat patterns of burglary. In addition, the police detection data indicate that pairs of detected burglaries occurring close in space and time are more likely to involve the same offender than distant pairs. Finally, the implications of the findings and limitations are discussed.

Weighted network Voronoi Diagrams for local spatial analysis

Detection of spatial clusters among geographic events in a planar space often fails in real world practices. For example, events in urban areas often occurred on or along streets. In those cases, objects and their movements were limited to the street network in the urban area. This deviated from what a set of freely located points could represent. Consequently, many of the spatial analytic tools would likely produce biased results. To reflect this limitation, we developed a new approach, weighted network Voronoi diagrams, to modeling spatial patterns of geographic events on street networks whose street segments can be weighted based on their roles in the events. Using kernel density estimation and local Moran’s index statistics, the frequency of events occurring on a street segment can be used to produce a weight to associate with the street segment. The weights can then be normalized using a predefined set of intervals. The constructed weighted Voronoi network explicitly takes into account the characteristics of how events distribute, instead of being limited to assessing the spatial distribution of events without considering how the structure of a street network may affect the distribution. This approach was elaborated in a case study of Wuhan City, China. Constructing weighted network Voronoi diagrams of these partitioned networks could assist city planners and providers of public/private services to better plan for network-constrained service areas.

Analyzing Urban Human Mobility Patterns through a Thematic Model at a Finer Scale

Taxi trajectories reflect human mobility over a road network. Pick-up and drop-off locations in different time periods represent origins and destinations of trips, respectively, demonstrating the spatiotemporal characteristics of human behavior. Each trip can be viewed as a displacement in the random walk model, and the distribution of extracted trips shows a distance decay effect. To identify the spatial similarity of trips at a finer scale, this paper investigates the distribution of trips through topic modeling techniques. Firstly, trip origins and trip destinations were identified from raw GPS data. Then, different trips were given semantic information, i.e., link identification numbers with a semantic enrichment process. Each taxi trajectory was composed of a series of trip destinations corresponding to the same taxi. Subsequently, each taxi trajectory was analogous to a document consisting of different words, and all taxi’s trajectories could be regarded as document corpora, enabling a semantic analysis of massive trip destinations. Finally, we obtained different trip destination topics reflecting the spatial similarity and regional property of human mobility through LDA topic model training. The effectiveness of this approach was illustrated by a case study using a large dataset of taxi trajectories collected from 2 to 8 June 2014 in Wuhan, China.

Characterizing street hierarchies through network analysis and large-scale taxi traffic flow: a case study of Wuhan, China

Hierarchy is an important property of a street network, which suggests that only a small number of streets are prominent. A previous empirical study of a European city has identified four levels of scale in a street network, namely the top 1%, top 20%, bottom 80%, and bottom 20%. This paper investigates such street hierarchies in a large Asian city, Wuhan, with a complicated network of streets. Based on network analysis, we find that street hierarchies in this case study are slightly different so that the fourth scale is adjusted from the initial 20 to 25%. The detected street hierarchies are further compared to the intensity of large-scale traffic flows at different time scales. We find that distributions of both daily and hourly traffic conform well to the street hierarchies. More specifically, the 20% of top streets accommodate about 98% of traffic flow, and the 1% of top streets account for more than 60% of traffic flow. Moreover, this finding indicates that the current street network of Wuhan needs to be improved because the top 20% of streets are rather overburdened leading to traffic congestion. Our study not only provides new quantitative evidence as to the emergence of street hierarchies but also highlights the possible traffic congestion.

The Local Colocation Patterns of Crime and Land-Use Features in Wuhan, China

Most studies of spatial colocation patterns of crime and land-use features in geographical information science and environmental criminology employ global measures, potentially obscuring spatial inhomogeneity. This study investigated the relationships of three types of crime with 22 types of land-use in Wuhan, China. First, global colocation patterns were examined. Then, local colocation patterns were examined based on the recently-proposed local colocation quotient, followed by a detailed comparison of the results. Different types of crimes were encouraged or discouraged by different types of land-use features with varying intensity, and the local colocation patterns demonstrated spatial inhomogeneity.

Urban Link Travel Time Prediction Based on a Gradient Boosting Method Considering Spatiotemporal Correlations

The prediction of travel times is challenging because of the sparseness of real-time traffic data and the intrinsic uncertainty of travel on congested urban road networks. We propose a new gradient–boosted regression tree method to accurately predict travel times. This model accounts for spatiotemporal correlations extracted from historical and real-time traffic data for adjacent and target links. This method can deliver high prediction accuracy by combining simple regression trees with poor performance. It corrects the error found in existing models for improved prediction accuracy. Our spatiotemporal gradient–boosted regression tree model was verified in experiments. The training data were obtained from big data reflecting historic traffic conditions collected by probe vehicles in Wuhan from January to May 2014. Real-time data were extracted from 11 weeks of GPS records collected in Wuhan from 5 May 2014 to 20 July 2014. Based on these data, we predicted link travel time for the period from 21 July 2014 to 25 July 2014. Experiments showed that our proposed spatiotemporal gradient–boosted regression tree model obtained better results than gradient boosting, random forest, or autoregressive integrated moving average approaches. Furthermore, these results indicate the advantages of our model for urban link travel time prediction.

Prediction of Suspect Location Based on Spatiotemporal Semantics

The prediction of suspect location enables proactive experiences for crime investigations and offers essential intelligence for crime prevention. However, existing studies have failed to capture the complex social location transition patterns of suspects and lack the capacity to address the issue of data sparsity. This paper proposes a novel location prediction model called CMoB (Crime Multi-order Bayes model) based on the spatiotemporal semantics to enhance the prediction performance. In particular, the model groups suspects with similar spatiotemporal semantics as one target suspect. Then, their mobility data are applied to estimate Markov transition probabilities of unobserved locations based on a KDE (kernel density estimating) smoothing method. Finally, by integrating the total transition probabilities, which are derived from the multi-order property of the Markov transition matrix, into a Bayesian-based formula, it is able to realize multi-step location prediction for the individual suspect. Experiments with the mobility dataset covering 210 suspects and their 18,754 location records from January to June 2012 in Wuhan City show that the proposed CMoB model significantly outperforms state-of-the-art algorithms for suspect location prediction in the context of data sparsity.

An open source toolkit for identifying comparative space-time research questions

Comparative space-time thinking lies at the heart of spatiotemporally integrated social sciences. The multiple dimensions and scales of socioeconomic dynamics pose numerous challenges for the application and evaluation of public policies in the comparative context. At the same time, social scientists have been slow to adopt and implement new spatiotemporally explicit methods of data analysis due to the lack of extensible software packages, which becomes a major impediment to the promotion of spatiotemporal thinking. The proposed framework will address this need by developing a set of research questions based on space-time-distributional features of socioeconomic datasets. The authors aim to develop, evaluate, and implement this framework in an open source toolkit to comprehensively quantify the changes and level of hidden variation of space-time datasets across scales and dimensions. Free access to the source code allows a broader community to incorporate additional advances in perspectives and methods, thus facilitating interdisciplinary collaboration. Being written in Python, it is entirely cross-platform, lowering transmission costs in research and education.

Analyzing Space-Time Dynamics of Theft Rates Using Exchange Mobility

A critical issue in the geography of crime is the quantitative analysis of the spatial distribution of crimes which usually changes over time. In this paper, we use the concept of exchange mobility across different time periods to determine the spatial distribution of the theft rate in the city of Wuhan, China, in 2016. To this end, we use a newly-developed spatial dynamic indicator, the Local Indicator of Mobility Association (LIMA), which can detect differences in the spatial distribution of theft rate rankings over time from a distributional dynamics perspective. Our results provide a scientific reference for the evaluation of the effects of crime prevention efforts and offer a decision-making tool to enhance the application of temporal and spatial analytical methods.