Situation awareness is a key requirement in managing civil contingencies, since major incidents, accidents and natural disasters are by their very nature highly unpredictable and confusing situations. It is important that those responsible for dealing with them have the best available information. The mash-up approach brings together information from multiple public and specialist sources to form a synoptic view, but the controller is still faced with multiple, partial and possibly conflicting reports from untrusted sources. The aim of this research is to investigate how the varying provenance of the data can be tracked and exploited to prioritise the information presented to a busy incident controller, and to synthesise a model or models of the situation that the evidence pertains to. The approach in this research is to develop a system involving novel approach and techniques to allow incident controllers and similar decision makers to augment official information input streams with information contributed by the wider public (either explicitly submitted to them or harvested from social networks such as Facebook and Twitter), and to be able to handle inconsistencies and uncertainty arising from the unreliability of such sources in a flexible way. The system takes in situational data in a structured format, such as the Tactical Situation Object (TSO) proposed by OASIS, a project funded by the European Framework Programme 6 (FP6) and performs an automated logical consistency checking in order to isolate inconsistent and absurd messages, identify the inconsistency between messages and cluster the consistent messages together. Each cluster of consistent messages that gives a possible view of a situation that the evidence pertains to is referred to as a `World View'. The logical consistency checking is performed using Alloy and Alloy Analyzer (sic). Finally, the system presents a set of possible world views, each internally consistent, which are ranked based upon an initial information provenance and quality metric (configured by the user) which is used to score the individual data items. The provenance and quality metric includes those factors that influence trust in information such as identity and location of informant, reputation, corroboration, freshness of information, etc. The result is a set of world views prioritised according to the provenance, trust and information quality metric. This thesis also presents some experimental results as proof of the concept. The experimentation has been carried out with a very small set of data to make the automation (automatic experimentation) feasible. However, a theoretical proof is offered to demonstrate the viability of the concept. Future work includes testing the system in real-life cases, in order to understand the utility of the system.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:658968 |
| Date | January 2014 |
| Creators | Rahman, Syed S. |
| Publisher | University of Warwick |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://wrap.warwick.ac.uk/71036/ |
Page generated in 0.0016 seconds