Data Collection and Representation Framework for Software and Human- Computer Interaction Measurements

摘要

This dissertation presents a framework for capturing, storing, and correlating human-computer interaction measures and software attribute measures. Discovering how computer programmers comprehend software has long been a problem in the field of computer science. Researchers want to understand what strategies programmers take as they look at software so that superior strategies can be investigated and documented. This involves collecting interaction and comprehension measures. Industry wants to determine when software will cause the greatest problems in comprehension, and if the software can be changed to ease this learning curve. This involves the collection of software attribute measures. A framework for the collection and correlation of both measures was developed based upon a straight-forward mathematical model. In this model, a user action is combined with the state of the software being manipulated at a specific point in time. This combination of action, software state, and time forms a unique tuple. This tuple is then the core element of analysis procedures which determine how software state affects user actions. From a mathematical standpoint, the framework components maintain sets of these tuples, organized by user session.