Qiuming Zhu

A multi-agent system of evidential reasoning for intelligence analyses

This paper describes a Multi-Agent System intended for assisting military commanders and intelligence analysts in the discovery and analysis of publicly available information that may have intelligence value (Open Source Intelligence, or OSINT). Our system is called Webster, which is a pun on the well-known dictionary and the World Wide Web. An innovative feature of Webster is the trust network that allows for the hierarchical integration of judgements provided by both human and computer agents, and the ability to extend the system by adding new agents that encapsulate a given characterization capability - such as the ability to provide a level of facial recognition on images that may be embedded in web pages. A key challenge is in creating a normalized concept structure or belief frame that all participating agents, at a certain level, can use to focus their analysis and render opinions that can be meaningfully combined with the opinions of other entities in the system. Webster can scale from a single machine to a large interconnection of subject matter experts and special-purpose computer systems by providing proxy agents that act as intermediaries in the system.

Collective agents interpolative integral (CAII) for asymmetric threat detection

This paper presents a reasoning system that pools the judgments from a set of inference agents with information from heterogeneous sources to generate a consensus opinion that reduces uncertainty and improves knowledge quality. The system, called Collective Agents Interpolation Integral (CAII), addresses a high level data fusion problem by combining, in a mathematically sound manner, multi-models of inference in knowledge intensive multi agent architecture. Two major issues are addressed in CAII. One is the ability of the inference mechanisms to deal with hybrid data inputs from multiple information sources and map the diverse data sets to a uniform representation in an objective space of reasoning and integration. The other is the ability of the system architecture to allow the continuous and discrete outputs of a diverse set of inference agents to interact, cooperate, and integrate.

Towards distributed ATR using subjective logic combination rules with a swarm of UAVs

In this paper, we present our initial findings demonstrating a cost-effective approach to Aided Target Recognition (ATR) employing a swarm of inexpensive Unmanned Aerial Vehicles (UAVs). We call our approach Distributed ATR (DATR). Our paper describes the utility of DATR for autonomous UAV operations, provides an overview of our methods, and the results of our initial simulation-based implementation and feasibility study. Our technology is aimed towards small and micro UAVs where platform restrictions allow only a modest quality camera and limited on-board computational capabilities. It is understood that an inexpensive sensor coupled with limited processing capability would be challenged in deriving a high probability of detection (Pd) while maintaining a low probability of false alarms (Pfa). Our hypothesis is that an evidential reasoning approach to fusing the observations of multiple UAVs observing approximately the same scene can raise the Pd and lower the Pfa sufficiently in order to provide a cost-effective ATR capability. This capability can lead to practical implementations of autonomous, coordinated, multi-UAV operations. In our system, the live video feed from a UAV is processed by a lightweight real-time ATR algorithm. This algorithm provides a set of possible classifications for each detected object over a possibility space defined by a set of exemplars. The classifications for each frame within a short observation interval (a few seconds) are used to generate a belief statement. Our system considers how many frames in the observation interval support each potential classification. A definable function transforms the observational data into a belief value. The belief value, or opinion, represents the UAVs belief that an object of the particular class exists in the area covered during the observation interval. The opinion is submitted as evidence in an evidential reasoning system. Opinions from observations over the same spatial area will have similar index values in the evidence cache. The evidential reasoning system combines observations of similar spatial indexes, discounting older observations based upon a parameterized information aging function. We employ Subjective Logic operations in the discounting and combination of opinions. The result is the consensus opinion from all observations that an object of a given class exists in a given region.

Epistemic Belief Frames in Distributed Effects-based Reasoning

Negotiation of a consistent ontology and concept label set is a significant problem in multi-agent systems (MAS). Heterogeneous MAS may not be able to fully negotiate a normalized ontology and so dealing with paraconsistent ontologies may be required. A belief calculus provides operators capable of expressing and maintaining uncertainty in paraconsistent reasoning operations over a belief frame. Developing and limiting a belief frame with heterogeneous ontologies is a particular problem for distributed MAS reasoning about causal chains. We propose using a sorted logic for entailing and populating epistemic belief frames supporting a belief calculus for distributed MAS effects-based reasoning

A Hybrid Cellular Inference Network for Multi-Agent System Organization

Multi-agent systems (MAS) must be able to ingest, correlate, and make decisions about disparate evidential sources. Our research on hybrid intrinsic cellular inference network (HICIN) provides a basis for developing reusable, partitionable, and distributable inference structures in an epistemic event-based space. The cellular regions of the inference space allow a gradation of belief measurements in a domain-specific hierarchy. Predicate annotations are encoded using WordNet-based ontology structure and standard synonym set labels, providing for disambiguation between subject domains while acting as a basis for domain specific partitioning of rule sets and belief regions.

Cellular inference network (CIN) for learning and control of multi-agent cooperation

The coordination and learning process in a multiagent system (MAS) can be modeled as a control and adaptation process of a hybrid dynamic system. We study an inference network organized in cellular structure for supporting multiple levels of abstraction and reasoning in learning and control of MAS. Specifically, we study how the cellular inference structure can serve as a representation of hybrid dynamic systems model and be applied to the knowledge gathering and refinement processes of the MAS. The study shows that it is a promising model for distributed reasoning in the presence of uncertainties in the domains of knowledge management, deconfliction, and integration.