Large Scale Behavioral Analytics via Topical Interaction

TL;DR

This paper introduces the split-diffuse (SD) algorithm that transforms dimension reduction outputs into uniformly distributed topic grids, enabling efficient visualization and analysis of large-scale textual data.

Contribution

The paper presents a novel SD algorithm that creates uniform topic grids from dimension reduction results, improving visualization and interaction with massive text datasets.

Findings

Efficient visualization of large-scale text data

Enhanced topical analysis and comparison

Improved data interaction capabilities

Abstract

We propose the split-diffuse (SD) algorithm that takes the output of an existing dimension reduction algorithm, and distributes the data points uniformly across the visualization space. The result, called the topic grids, is a set of grids on various topics which are generated from the free-form text content of any domain of interest. The topic grids efficiently utilizes the visualization space to provide visual summaries for massive data. Topical analysis, comparison and interaction can be performed on the topic grids in a more perceivable way.