Structured Data Analysis: A Cognition-Based Design for Data Analysis Software
NORTH CAROLINA UNIV AT CHAPEL HILL DEPT OF COMPUTER SCIENCE
Pagination or Media Count:
We present a structured data analysis environment designed to improve data analysts productivity over present unstructured environments. The environment is based on three fundamental concepts Cognitive Modes, Dataflow Diagrams, and Language Generation. We hypothesize that data analysts adopt different cognitive modes of behavior and thought for different stages of the data analysis process. Our environment has corresponding system modes that visually represent these cognitive modes. One of the system modes enables users to construct dataflow diagrams by manipulating icons representing datasets, data analysis procedures, and data analysis results. In this mode the user structures the data analysis by drawing lines from dataset icons to procedure icons the computer in turn performs the analysis defined by the dataflow diagram, display result icons and draws lines from the procedure icons to the results icons to complete the diagram. Then, the icons and lines form a manipulable dataflow diagram which represents the flow of data from a dataset through an analysis procedure into results.
- Computer Programming and Software