AI-powered classroom discourse analysis tool that visualizes the structure of classroom discussions.
AIを活用して授業記録を分析し、発言間の関係性、トピック構造、隠れた影響者を可視化するツールです。
- Relation Network / 関係ネットワーク: Visualize relationships between utterances (agreement, disagreement, addition, etc.)
- Statistical Analysis / 統計分析: Speaker statistics, centrality analysis, topic distribution
- Timeline View / タイムライン表示: Chronological view of all utterances with coding
- Bilingual Support / 多言語対応: Japanese and English interface
- Export / エクスポート: Export analysis results as JSON, charts as PNG/JPG
- Node.js (v18 or later) - Download
- Gemini API Key - Get it from Google AI Studio
- Install Node.js
- Download and extract this project
- Double-click
start-windows.bat - Open http://localhost:3000 in your browser
- Install Node.js
- Download and extract this project
- Open Terminal in the project folder
- Run:
chmod +x start-mac-linux.sh ./start-mac-linux.sh
- Open http://localhost:3000 in your browser
# Install pnpm (if not installed)
npm install -g pnpm
# Install dependencies
pnpm install
# Start development server
pnpm devThen open http://localhost:3000 in your browser.
The CSV file should have 3 columns: CSVファイルは3列で構成されます:
発言番号,発言者,発言内容
1,T,今日は三角形について学びます
2,A子,先生、三角形って何ですか?
3,T,いい質問ですね。三角形とは...
- Column 1: Utterance number / 発言番号
- Column 2: Speaker name (T, 先生, 教師 = Teacher) / 発言者名
- Column 3: Utterance content / 発言内容
- Enter your Gemini API key / Gemini APIキーを入力
- Upload a CSV file / CSVファイルをアップロード
- Wait for AI analysis / AI分析を待つ
- Explore the results in different tabs / 各タブで結果を確認
The easiest way to deploy is using Vercel:
- Push to GitHub
- Connect to Vercel
- Deploy automatically
# Build for production
pnpm build
# Start production server
pnpm startMIT License
For issues or questions, please create an issue on GitHub. 問題や質問がある場合は、GitHubでissueを作成してください。