Supporting Intelligence Analysts with a Trust-Based Question-Answering System

Abstract

Intelligence analysts have to work in highly demanding circumstances. This causes mistakes with severe consequences, which is the reason that support systems for intelligence analysts have been developed. The support system proposed in this paper assists humans by offering support that improves their performance, without reducing them in their freedom. This is done with a trust-based question answering system (T-QAS). An important part of T-QAS are trust models which keep track of trust in each of the agents gathering information. Using these trust models, the system can support the intelligence analyst by: 1) helping to decide which agents are trusted enough to receive questions, 2) providing information about the reliability of each of the sources used, and 3) advising in making decisions based on information from possibly unreliable sources. An implementation of last two capabilities of T-QAS is evaluated in an experiment in which participants perform a decision making task with information from possibly unreliable sources. Results show that the proposed T-QAS support indeed helps participants to improve their performance. We therefore expect that future intelligence analyst support systems can benefit from the inclusion of T-QAS.