A Proposal for Semantic Integration of Crime Data in Mexico City

Crime is a common problem in big cities where the authorities regularly update data crime reports. In Mexico City, the crime reports are available as open data. However, other relevant data are not connected to them (e.g., socioeconomic data). Therefore, the socioeconomic and geographic data can help understand how the crime is characterized and what social indicators are related to it. In this research, we explore how data crime reports are described and how they can be associated in an Ontology with other data, such as socioeconomic and geographic data. The goal is to discover the social indicators related to a particular crime in a specific area by using SPARQL queries from a knowledge representation. Then, data sets from crime reports, socioeconomic and geographic data from 2016 were integrated to explore crime behavior in Mexico City. The work uses a NeOn methodology in which resources from existing ontologies or non-ontological resources can be mixed. Next, a set of SPARQL queries is defined to extract the knowledge from ontology and discover the associations between crime in geographic and socioeconomic domains. The results showed a set of queries where it is possible to know where a crime occurred and what other factors are associated with the crime and help to identify possible patterns among them.