Algorithms that Empower? ACS and the Future of Intelligence Analysis

• In Event: The Social Life of Algorithms - II

Abstract

Although intelligence analysts have access to an increasingly wide variety of analytic tools and tradecraft they can draw on, applying these to a problem can require a great deal of effort for the analyst to: (1) learn about the tool/method and how to effectively use it on a given problem; and (2) adapt the analyst’s workflow to efficiently incorporate it. A new kind of computational architecture, called the Analytic Component System (ACS), developed by the National Security Agency (NSA)-funded Laboratory for Analytic Sciences, aims to address these challenges. The ACS platform includes variety of analytic components that might include automated algorithms, interactive tools, or manual techniques. One of the main objectives of the Analytic Component System is to align the various algorithms, tools, and techniques with an explicit framework that enables the capture, analysis, and support of the decisions analysts make in the development of their analytic workflow. Capturing these decisions can allow creators, consumers, and reviewers of the analytic products to assess the rigor of the workflow that produced it. Creating a repository of these workflows can provide an invaluable data source to methods that seek to recommend activities and workflows for other analytic problems. This kind of computational architecture is designed to empower intelligence analysts, who work in a stressful, time-pressured, “Big Data,” and high stakes work environment. This paper will discuss interviews conducted with the ACS design team and aim to situate the work in the context of STS literature on user- and values-centered computer design.

Author

• Kathleen Vogel, Prof.