Supporting Representation Management in Intelligence Analysis through Automated Decision Aids

Abstract

Intelligence analysts-overloaded with complex and disparate data-must incorporate information into cohesive and convincing narratives and explanations. For analysts, data overload can result in developing premature conclusions and limit their ability to effectively conduct comprehensive analyses. Historically, automated decision aids designed to help with these processes have largely failed analysts in managing a fundamental work tradeoff between analytic narrowing and broadening because software tools too often attempt to supplant analyst reasoning, rather than support the iterative process as a whole. Instead of such brittle approaches, previous research suggests analysts would benefit from automation that supports a collaborative partnership throughout the analysis process. The research and development described in this work began with a cognitive analysis of practicing analysts, entailing literature review, interviews, and walkthroughs with a software prototype. Findings from the cognitive analysis were translated into design requirements for decision aids that support representation management across multiple intelligence products. The findings provide insight into under-supported cognitive and collaborative work in modern intelligence analysis with implications for the future design of useful automated decision aids.

Analysis and Requirement Generation for Defense Intelligence Search: Addressing Data Overload through Human–AI Agent System Design for Ambient Awareness

“…This model is structured with three abstract functions: "Down-Collect", "Conflict & Corroboration", and "Hypothesis Exploration", and it incorporates functional requirements for analysis with requirements for effective teaming. Recent work by Voshell et al, extending the work of Elm and others, emphasizes the challenges analysts face in managing data overload and the limitations of existing automated decision aids [21]. Their research suggests that analytic support tools have historically failed because they too often try to replace analyst reasoning instead of supporting the iterative process.…”

A model of symbiomemesis: machine education and communication as pillars for human-autonomy symbiosis

“…The former allows mutual understanding of the messages, while the latter allows mutual understanding of the rationales behind actions and decisions. Interpretability, explainability and predictability form the tenets for transparency [14] and augment the human user with a better ability to direct/re-direct AIAS as priorities change [15]…”

Combining cognitive engineering and information fusion architectures to build effective joint systems

“…DNDW combines the structural elements of the dual node network (DNN) technical architecture for data fusion and resources management [11] with the cognitive engineering elements of decision ladders [12], [13], [14], [15] and decision wheels [16] , helping alleviate both perennial challenges in intelligence analysis as well as new "big data" challenges [2] through more fluid coordination of support technologies within organizational and cognitive decision-making processes.…”

“…As intelligence analysts work with automated tools to understand collection requirements, exploit and analyze new multi-INT collection, navigate expansive databases, and collaborate across distributed work environments , they face many challenges that extend beyond the barriers of individual problem solving. One of the most challenging aspects for introducing new support systems for effective information analysis involves closing the gap between the increasing amounts of data availability and the lack of useful automation to support human analysis of this information [1], [2] . Elm et al's [3] hypothesized support model of information analysis-extended by Roth et al [4] -serves as an evolving framework to identify information needs and decision-making requirements for designing effective intelligence analysis tools and systems .…”