Table 3 An overview of the 32 selected papers with information about the space and time of the research, the crime data, and forecasting details
No* | Authors and date | Space | Time | Crime Data | Forecasting | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
Study area | Scale | Sampling period | Months | Type | Sample | Inference | Task | Spatial unit | Temporal unit | ||
1 | Araujo Junior et al. (2017) | Natal, Brazil | City | 2006–2016 | 132 | U | U | # of crimes | Regression | Rectangular grid (U), districts | Week |
2 | Araújo et al. (2018) | Natal, Brazil | City | 2006–2016 | 132 | U | U | Hotspots | Binary classification | k-means cells of varying size (U) | Week |
3 | Bowen et al. (2018) | DeKalb, USA | County | 2011–2014 | 48 | Violent crime | U | Hotspots | Binary classification | Census block groups | Month |
4 | Brown and Oxford (2001) | Richmond, USA | City | 1994–1999 | 72 | Break and enter | ≈ over 24,000 | # of crimes | Regression | Grid cells of 1.66 km2, precincts | Week, month |
5 | Cohen et al. (2007) | Pittsburgh, USA | City | 1991–1998 | 96 | 2 crime types | 1.3 million | # of crimes | Regression | 1219 m × 1219 m grid cells | Month |
6 | Dash et al. (2018) | Chicago, USA | City | 2011–2015 | 60 | 34 crime types | 6.6 million | # of crimes | regression | Communities | Month, year |
7 | Drawve et al. (2016) | Little Rock, USA | City | 2008–2009 | 18 | Gun crime | 1429 | Hotspots | Binary classification | 91 m × 91 m grid cells | 6 months |
8 | Dugato et al. (2018) | Milan, Italy | City | 2012–2014 | 36 | Residential burglary | 20,921 | Hotspots | Binary classification | Grid cells of 2500 m2 | Year |
9 | Gimenez-Santana et al. (2018) | Bogota, Colombia | city | 2012–2013 | 24 | 3 crime types | U | Hotspots | Binary classification | 75 m × 75 m grid cells | Year |
10 | Gorr et al. (2003) | Pittsburgh, USA | City | 1991–1998 | 96 | 5 crime types | ≈ 1 million | # of crimes | Regression | Police precincts | Month |
11 | Hart and Zandbergen (2014) | Arlington, USA | City | 2007–2008 | 24 | 4 crime types | 6295 | Hotspots | Binary classification | Grid cells of 3 different sizes (U) | Year |
12 | Hu et al. (2018) | Baton Rouge, USA | City | 2011 | 12 | Residential burglary | 3706 | Hotspots | Binary classification | 100 m × 100 m grid cells | Week |
13 | Huang et al. (2018) | New York, USA | City | 2014 | 12 | 4 crime types | 103,913 | Category of crime | Binary classification | Districts | Day, month |
14 | Ivaha et al. (2007) | Cardiff, UK | City | 2001–2003 | 26 | Criminal damage | U | Percent of crime in clusters | Regression | Clusters of varying size (U) | Day |
15 | Johansson et al. (2015) | Sweden three cities: Stockholm, Gothenburg, and Malmö | Cities | 2013–2014 | 12 | Residential burglary | 5681 | Hotspots | binary Classification | Grid cells (U) | 3 months |
16 | Kadar and Pletikosa (2018) | New York, USA | City | 2014–2015 | 24 | All crime and 5 crime types | 174,682 | # of crimes | Regression | Census tract | Year |
17 | Liesenfeld et al. (2017) | Pittsburgh, USA | City | 2008–2013 | 72 | All crime | 9936 | # of crimes | Regression | Census tracts | Month, year |
18 | Lin et al. (2018) | Taoyuan City, Taiwan | City | 2015–2018 | 39 | Motor vehicle thefts | ≈ 8580 | Hotspots | Binary classification | 5 to 100 × 5 to 100 grid cells (U) | Month |
19 | Malik et al. (2014) | Tippecanoe, USA | County | 2004–2014 | 120 | all crime | ≈ 310,000 | Hotspots | Binary classification | Grid cells (U), law beats, census blocks | Week |
20 | Mohler (2014) | Chicago, USA | City | 2007–2012 | 72 | 2 crime types | 78,852 | Hotspots | Binary classification | 75 m × 75 m, 150 m × 150 m grid cells | Day |
21 | Mohler and Porter (2018) | Portland, USA | City | 2012–2017 | 60 | 4 crime types | U | Hotspots | Binary Classification | Grid cells of 5806 m2 | Week, 2 weeks, month, 2 months, 3 months |
22 | Mohler et al. (2018) | Indianapolis, USA | City | 2012–2013 | 24 | 4 crime types | U | Hotspots | Binary classification | 300 m × 300 m grid cells | Day |
23 | Mu et al. (2011) | Boston, USA | City | 2006–2007 | 24 | Residential burglary | U | Hotspots | Binary classification | 20 × 20 grid cells (U) | Month |
24 | Rodríguez et al. (2017) | San Francisco, USA | City | 2003–2013 | 120 | Burglary | U | Properties of clusters | Regression | Clusters (U) | Day |
25 | Rosser et al. (2017) | “Major city”, UK (U) | City | 2013–2014 | 13 | Residential burglary | 5862 | Hotspots | Binary classification | Street segments (U) | Day |
26 | Rumi et al. (2018) | Brisbane, Australia; New York City, USA | Cities | 2013–2013 (AUS); 2012–2013 (USA) | 9 and 12 | 6 crime types | U | Hotspots | Binary classification | Census regions | 3 h |
27 | Rummens et al. (2017) | “Large city”, Belgium (U) | city | 2011–2014 | 48 | 3 crime types | 163,800 | Hotspots | Binary classification | 200 m by 200 m grid cells | 2 weeks, daytime month, night time month |
28 | Shoesmith (2013) | USA | Country | 1960–2009 | 600 | 2 crime types | U | Crime rate | Regression | USA regions | Year |
29 | Yang et al. (2018) | New York, USA | city | January 2014–April 2015 | 16 | 7 crime types | U | Hotspots | Binary classification | 0.01 latitude × 0.01 longitude grid cell size | Day, week, month |
30 | Yu et al. (2011) | “City in the Northeast”, USA (U) | City | U | U | Residential burglary | U | Hotspots | binary Classification | grid cells (U) | Month |
31 | Zhao and Tang (2017) | New York, USA | City | 2012–2013 | 12 | U | U | # of crimes | Regression | 2 km × 2 km grid cells | Day, week |
32 | Zhuang et al. (2017) | Portland, USA | City | March 2012–December 2016 | 58 | All crime | U | Hotspots | Binary classification | 183 m × 183 m grid cells | 2 weeks |