Article Preview
TopBackground
Gruenwald et al. propose an ontology to improve the search in a database system (Gruenwald et al., 2003). In their system, they leverage the reasoning power of the ontological system to assist in the creation and amendment of a counter-terrorism database. Turner et al. also proposes the use of an ontology for the analysis of terrorist attacks (Turner et al., 2011), however, there are some key concepts for counter-terrorism that are missing such as ideology and modus operandi; it is also unclear what the exhaustive taxonomy for weapons is. Mannes and Golbeck offer practical advice about how to construct a useful ontology for the terrorism domain (Mannes & Golbeck, 2007). Other researchers (Inyaem et al., 2009) use an ontology as a named entity feature selection technique to be used as a basis for machine learning algorithms to extract information related to terrorism events from Thai news articles.
More recently, in 2019, Sheremetyeva and Zinoveva (Sheremetyeva & Zinoveva, 2019) used an ontology to analyse e-news for the terrorism domain. This work is a good starting point to model the domain. The reliance on e-news data does restrict the ontological model. Therefore, for the purpose of this work, the terrorism event was not able to be adequately represented using other classifications such as arms, arson, etc.
Most recently, researchers (Jindal et al., 2020; Rawat, 2022) propose a semi-automatic way for constructing domain ontologies using the terrorism domain. These works perform the population of an ontology using a combination of formal concept analysis (a mathematical strategy) and social networks i.e., Twitter feeds which is different to using an OBDA approach. The researchers’ method could be a useful approach for populating instance level data for an ontology, but the semantic relations concerning the domain still needs to be represented and refined.