SEMANTICAL INTEROPERABILITY FOR ENTERPRISE DECISION SUPPORT SYSTEM IN ACTIVE ENVIRONMENT

Abstract

Semantics are playing key role in the area of decision making. Besides playing an importance role in web, semantics-based decision making is getting ingress in other fields, especially in war related technologies. Our proposed framework for rescue 1122 Pakistan, is also uses semantics and artificial intelligence combined, to enhance the decision making process and create situation awareness. A simple plane English string defining the incident works as input to our framework, from that string framework extracts contact information with the help of regular expression and extracts address information with the help of knowledge base and semantics. The rest of the string goes into lexical analyzer, which expands this string, converts it into new sets of ontologies, neglects/discards connecting words and parses new relationships inside newly formed sets of ontologies with the help of relationship builder. Newly formulated strings then are used to create new sets of semantical senses by implementing open source coreNLP and WordNet library rules. Each set of semantical senses, then represented in equally distinct numerical values for processing to ID3 algorithm implemented in Accord.Net library. This algorithm uses multiple iterations of Iterative Dichotomiser machine learning algorithm to predict results with a much more accuracy. Finally, decision are taken regarding resources to be utilized in the rescue operation, location of accident and victims, contact details of accident reporting person, incident type and related rescue activities to be carried out. This guides the rescue operation till hospitalization of the victim if required, with the help of geographical tagging of the location over the map inside rescue vehicle and control center’s common operating picture.