Crime Intelligence from Social Media Using - development and representation of an ontology
MetadataShow full item record
Full text not available
Kriminelle bruker mer og mer sosiale nettverk til sine fordeler. Derfor har politietproblemer med å håndtere den store mengden data. De ekstraherte dataene frasociale nettverk er ustrukturerte, noe som gjør det utfordrende å filtrere ut nød-vendig informasjon. Denne oppgaven utvider SMONT-ontologien for å lukke dettegapet. Den utvidede versjonen heter "Crime Intelligence from Social Media" (CIS-MOv2) og samler digital bevis fra sosiale nettverkThese days, criminals take advantage of Online Social Networks (ONS) in many different ways. Therefore it can be defined as a central point of cybercrime. Legal enforcement agencies (LEAs) have trouble dealing with the large amount of data from OSNs. The extracted data from OSNs is unstructured, making it challenging for LEAs to filter out needed intelligence, which can be used in the legal domain. There is no ontology model, after reviewing the literature, that explains all points of crime investigation with the aim of data integration, information sharing, collection, query answering as well as the preservation of digital evidence with the help of biometric features. This paper aims to develop and present an extended version of the SMONT ontology to close this gap. The extended version is called "Crime Intelligence from Social Media" (CISMOv2) and gathers digital evidence from OSNs, so LEAs can create investigative systems to antagonize cybercrime. The CISMO ontology presents the core concepts that correlate with criminal cases in the police repositories, biometric data, and OSNs evidence. So it likely for LEAs to clear up crimes, put light on crime patterns, and construct future crime patterns. CISMO is more concise and contains a broader concept knowledge base compared to the SMONT ontology. Case studies of general topics of criminal cases in OSNs help to verify the effectiveness of CISMO. They show how the semantic approach can help LEAs collect knowledge for crime investigation by using natural language processing and machine learning.