Towards a Process for Criminal Semantic Information Fusion to Obtain Situational Projections

Abstract

Situational Awareness (SAW) refers to the level of consciousness that a human holds about a situation. In risk management domain, SAW failures can induce human to make mistakes in decision making. In addition, criminal domains with dynamic situations are prone to information quality problems, especially when they are provided by humans. Considering the nature of the information and the context, the information may be incomplete, outdated, inconsistent or influenced by cultural and stress factors. Other limiting factors are related to the ability to deal with large scale data, hindering informational processes such as processing, storage and retrieval of information. Information fusion processes present opportunities to improve the quality of information, generating subsidies that can contribute to a more complete SAW. The state-of-the-art presents solutions that involve the representation and processing of high-level information, however applying fusion techniques that are limited to the analysis and integration of information, where the application of semantics, ontological models and the concern with the information quality is limited. The proposal of this work is the development of a semantic information fusion, able to generate better quality information, aiming to make situational projections. Moreover, the new fusion process, as an extension of a previous human-driven fusion model, handles an application ontology, able to represent situations of the risk management domain and enable semantic inferences. Results so far validate the need of semantic-based fusion approaches for the development of useful risk assessment solutions, both to enhance SAW and empower critical decision-making.