Joint Human-Machine Data Exploration Sandbox

Information

• Publication Type: Student Project
• Workgroup(s)/Project(s):
  • Joint Human-Machine Data Exploration
• Date: 2024
• Date (Start): September 2023
• Date (End): September 2024
• Matrikelnummer: 00925239
• First Supervisor: Manuela Waldner

Abstract

Data analysis exploration is becoming increasingly challenging as datasets grow in scale and complexity. The Joint Human-Machine Data Exploration (JDE) framework offers a novel solution for analyzing large, unstructured datasets by integrating human insight with machine learning. The framework facilitates dynamic user interaction and visual exploration through three interconnected views: the data view, frame view, and knowledge view. These views enable users to align data exploration with evolving knowledge models. Implemented as a flexible, modular system using a client-server architecture, the JDE framework supports interactive data manipulation and real-time feedback. This project developed a functional prototype, the JDE sandbox, showcasing the system's potential for enhancing exploratory data analysis. Future work will focus on expanding the framework's capabilities and improving accessibility for a broader audience.

Additional Files and Images

Weblinks

• GitLab
• demo
  Online demo of the sandbox

BibTeX

@studentproject{wolf-2024-jhd,
  title =      "Joint Human-Machine Data Exploration Sandbox",
  author =     "Dominik Wolf",
  year =       "2024",
  abstract =   "Data analysis exploration is becoming increasingly
               challenging as datasets grow in scale and complexity. The
               Joint Human-Machine Data Exploration (JDE) framework offers
               a novel solution for analyzing large, unstructured datasets
               by integrating human insight with machine learning. The
               framework facilitates dynamic user interaction and visual
               exploration through three interconnected views: the data
               view, frame view, and knowledge view. These views enable
               users to align data exploration with evolving knowledge
               models. Implemented as a flexible, modular system using a
               client-server architecture, the JDE framework supports
               interactive data manipulation and real-time feedback. This
               project developed a functional prototype, the JDE sandbox,
               showcasing the system's potential for enhancing exploratory
               data analysis. Future work will focus on expanding the
               framework's capabilities and improving accessibility for a
               broader audience.",
  month =      sep,
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2024/wolf-2024-jhd/",
}