Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
An intelligent tool for experienced programmers learning Ada
Abstract
When experienced programmers learn a new language, their previous planning knowledge results in both positive and negative transfer to the new language. A tool to help them learn the new language must include planning as an important component and the planning help must consider the previous planning knowledge of the programmer. Several existing programming tutors guide novice users through a few simple problems and spend at least as much time on coding as on planning. This dissertation describes an intelligent tool built to help experienced Pascal or C programmers learn to use Ada packages to create reusable software components. Empirical studies were completed to determine the knowledge base of correct and buggy plans used by programmers as they create modules which use packages. This knowledge base is used to model the student and to diagnose errors. Given an exercise, the student creates a solution by working from high-level plans to more detailed plans to code. This method of solution simulates a programmer using pseudocode for planning and replacing plans with code when the way to implement a plan becomes clear. When it is clear that the student is using a plan from a previous language that is impossible or extremely suboptimal in Ada, the tool intervenes to get the user back on a productive path. The tool uses model-tracing for diagnosis. In some situations, students choose from a menu to indicate their mental state, and in other situations, plan recognition is used to determine their mental state from their intermediate actions. The initial version of the tool was evaluated in a study in which programmers at the senior or graduate level used the tool to solve one problem and then completed transfer tasks. The results of the study showed that subjects learned package concepts using the tool. The results also suggest improvements for the next version of the tool.
Subject Area
Computer science|Artificial intelligence
Recommended Citation
Fix, Vikki Lorraine, "An intelligent tool for experienced programmers learning Ada" (1993). ETD collection for University of Nebraska-Lincoln. AAI9322795.
https://digitalcommons.unl.edu/dissertations/AAI9322795