Improving Visualization and User Interaction in Histree

Jupyter Lab is a commonly used programming environment, especially in the field of data science and machine learning. It provides an intuitive user interface with the ability to execute code, compare the content of different notebooks and even visualize data sets. This thesis focuses on enhancing the existing Histree plugin, which visualizes notebook experiment history in Jupyter Lab as a tree structure and provides interactive options for personalized visualization. Despite its utility, the plugin currently faces challenges. Feedback showed that the plugin struggles with the comprehensive visualization of large trees and is missing some interaction methods to further personalize the visualization. To address these issues, this work introduces new visualization and interaction methods aimed at enhancing the user experience with the plugin, such as a listing visualization, a time based slider and an annotation feature. These features were selected based on user interface design principles. They were evaluated through a user feedback study, and based on the feedback received, are refined and implemented into the existing Histree plugin.

Project information

Status:

Finished

Thesis for degree:

Bachelor

Student:

Maryan Vogel

Supervisor:
Part of research project:

SE4ML - Processes, People and Tools

Id:

2024-009