Cognitive Task Analysis of Data Mining Processes in Bioinformatics Research

Abstract

Health care processes generate vast amounts of data of potential value to improve health and health care services. Bio-medical data, such as genetic data, need to be properly stored in data warehouses and analyzed to benefit health care. From a multidisciplinary perspective, Bioinformatics addresses research, development and application of computational tools and approaches for analyzing and thus expanding the use of biological, medical, behavioral and health data. Data mining techniques are often used by bioinformatics researchers. In the current project, a bootstrapping-themed Cognitive Task Analysis was adopted to analyze the data mining processes used by several bioinformatics researchers at the University at Buffalo, SUNY. Two of the Applied Cognitive Task Analysis (ACTA) techniques, “Task Diagram Interview” and “Knowledge Audit Interview,” were integrated with questionnaire surveys and field observations. This paper presents the ACTA portion of the cognitive task analysis, and shows that there was agreement on most of the cognitive elements required for data mining.