Keywords

Overview

Keywords in Mantis are auto-extracted or user-defined terms that help describe, group, and filter data points. They act as thematic tags and anchors for exploration, especially in text-heavy datasets.

Purpose

• Let you quickly identify what a cluster or point is about.
• Enable search, color-coding, and organization based on semantic similarity.
• Drive insights through keyword merging, grouping, and summarization.

Key Features

• Auto-Extraction: Mantis agents extract keywords from documents, summaries, or clusters.
• Spider View: Visualize keyword relationships with draggable webs showing which points relate to each term.
• Keyword Merging: Related terms can be grouped using similarity logic (e.g., “governance” and “policy”).
• Color Coding: Keywords are assigned pastel colors, which also highlight related points on the map.

Tips

• Use keywords to filter or search with terms like "Keyword search: climate, energy".
• Combine keyword overlays with bags or clustering for multi-layered insights.
• Use merged keywords to reduce clutter in dense thematic spaces.