📌 Project Overview

While computational models can segment Georgian word forms (see this related project), human researchers and learners need a way to visualize those segments to identify patterns. This project is a custom-built web dashboard designed to take complex morphological data and present it in a highly structured, comparative format.

⚡ Live Demo

See the live, interactive demo here.

⚙️ Core Features & Linguistic UX

The interface is engineered specifically for the unique challenges of Kartvelian morphology:

  • Synchronized Morphological Alignment: Verb charts are structured into a master grid that vertically aligns every grammatical slot—from preverbs to roots—allowing for instant visual comparison across different forms.
  • Interactive Cross-Form Highlighting: Hovering over a specific morpheme automatically illuminates its counterparts across the entire paradigm, making it easy to trace individual markers throughout complex conjugation cycles.
  • Visual Stem & Series Tracking:
    • Automated Diffs: The system detects and color-codes variations between different verb series (e.g., Present vs. Future), clearly isolating stem developments and thematic changes.
    • Grammatical Anchoring: Each screeve (grammatical category) is visually distinct, using unique accents to help users stay oriented within the larger verb system.
  • Live Data Integration: Powered by a Google Apps Script backend, the dashboard dynamically pulls from a relational database that links inflected word forms back to their base lemmas.

🛠 Technical Stack

  • Frontend Framework: Vue.js for reactive state management and search functionality.
  • Styling: Advanced CSS Grid and custom properties for character-perfect vertical alignment.
  • Backend: Google Apps Script (GAS) acting as a lightweight API and host.
  • Typography: Optimized for the Mkhedruli script using system fonts for maximum legibility.

🔗 The Data Engine

The segmentation data visualized here is powered by a CRF-based machine learning model.

← Back to Projects