CommentVis: Unveiling Comment Insights Through Interactive Visualization Tool

Recently, online forums have emerged as a stage for consumers to comment and share their reviews. These user comments serve as a valuable data source for marketing professionals and analysts. Nevertheless, conventional user interfaces often present an overwhelming volume of comments in a linear-stru...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE Pacific Visualization Symposium S. 353 - 358
Hauptverfasser: Yan, Guangjing, Jang, Jinhwa, Seo, Jinwook
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 23.04.2024
Schlagworte:
Computing methodologies-Artificial intelligence-Natural language processing
Data processing
Data visualization
Feature extraction
Human-centered computing-Visualization-Visualization systems and tools
Reviews
Semantics
Skeleton
Soft sensors
ISSN:2165-8773
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Recently, online forums have emerged as a stage for consumers to comment and share their reviews. These user comments serve as a valuable data source for marketing professionals and analysts. Nevertheless, conventional user interfaces often present an overwhelming volume of comments in a linear-structured list, significantly impeding the efficiency of marketing professionals in analyzing the feedback. In response to this challenge, we introduce CommentVis, an interactive visualization tool that helps users grasp the skeleton of comments' semantic distribution. Using the tool, analysts gain detailed information about comments with profound insights. The tool leverages state-of-the-art large-scale language models, optimizing the speed and depth of analysis of substantial text data that may take a long time for marketers. To illustrate the practical application of CommentVis, we present a usage scenario that demonstrates its effectiveness in real-world marketing analysis. The tool's impact and utility were further validated through a user study involving three marketing professionals in a global manufacturing company.
ISSN:2165-8773
DOI:10.1109/PacificVis60374.2024.00050