Willow : a tool for interactive data structures and algorithms visualization

• Autor(a) principal: MORAES, Pedro Henrique Sousa de
• Data de Publicação: 2020
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Institucional da UFPE
• Texto Completo: https://repositorio.ufpe.br/handle/123456789/39283
• topic: Engenharia de software; Linguagem de programação

Teaching Introductory Programming and Data Structures and Algorithms is an important part of Information Technology courses. Both disciplines include essential concepts for software development. Preparing lessons for these courses can be time demanding and tedious as instructors often need to create and modify examples using slides and sketches on a board. Students may also have difficulties due to the high level of abstraction of the content taught in both courses. Educational visualization tools, such as Python Tutor exist, but they provide rigid choices of visualization schemes used to represent the data. Most educational tools are discontinued or have limited support to the visualization of data structures and algorithms. Other tools create visualizations of several algorithms, but lack the ability to edit the source code or inputs. This work proposes Willow, a web-based interactive tool to visualize program state. Willow enables the user to customize visualizations and to walk through the code in both directions to facilitate code understanding. The sensible features of Willow are its ability to change data representations, jump to any point of a program with visual support during debug sessions, and detection and animation of common data structures such as lists and trees. To evaluate Willow, we conducted two studies, a survey with instructors of several universities, and a follow up experiment with programmers of a freelancing platform. We obtained positive feedback from 91% of the survey participants, suggesting that Willow can be used as an teaching aid tool by instructors. In the follow up experiment with programmers, we could not find significant difference between participants that used Willow and participants that did not, the results of the experiment were not conclusive. Nevertheless, we obtained positive results after considering a subset of the experiment tasks, participants also reacted positively to the tool and many would like to use it again.