README.md

metadata

title: DASS QuickText Model
emoji: 📚
colorFrom: indigo
colorTo: red
sdk: gradio
sdk_version: 6.8.0
app_file: app.py
pinned: false

🌿 心理健康風險程度測試 (DASS-12)

這是一個基於 DASS (Depression, Anxiety and Stress Scale) 簡化版開發的自動化心理健康評估工具。透過 12 題心理狀態問答，利用機器學習模型預測潛在的心理健康風險程度。

🚀 功能特點

• 即時評估：透過互動式介面，5 分鐘內獲得評估結果。
• 三維度分析：分別計算「焦慮」、「憂鬱」與「壓力」的分數比重。
• 雲端紀錄：測試結果會自動去識別化並加密儲存至後端雲端資料庫（Google Sheets），供學術研究與趨勢分析。
• 歷史紀錄：支援瀏覽器當次 Session 的歷史紀錄查看。

🛠 技術棧

• 前端介面: Gradio
• 數據處理: Pandas, Numpy
• 機器學習: LightGBM (LGBMClassifier)
• 後端儲存: Google Sheets API (via gspread)
• 部署環境: Hugging Face Spaces

📁 檔案說明

• app.py: Gradio 主程式與邏輯控制。
• DASS_model.bin: 已訓練完成的機器學習模型。
• AutoPreprocess.py: 自定義數據預處理類別。
• requirements.txt: 專案執行環境所需的 Python 套件清單。

🔐 隱私聲明

本專案極為重視使用者隱私：

1. 使用者輸入之「暱稱」僅用於介面呈現，不會被傳送至雲端資料庫，亦不會用於模型計算。
2. 儲存之數據包含：測試時間、人口學基本資訊（性別/年齡/家庭人數）、各維度分數及 12 題原始答案。
3. 本工具非醫療器材，結果僅供參考。

👥 開發團隊

• 第四組：心理健康風險預測小組 🌿

• 🌿 Psychological Health Risk Assessment (DASS-12)

This is an automated psychological health evaluation tool based on the simplified DASS (Depression, Anxiety and Stress Scale). It leverages a Machine Learning model to predict potential mental health risk levels based on 12 psychological state questions.

🚀 Key Features

• Instant Assessment: Interactive interface that provides evaluation results within 5 minutes.
• Three-Dimensional Analysis: Calculates and visualizes specific scores for Anxiety, Depression, and Stress.
• Cloud-Based Storage: Test results are anonymized and securely stored in a cloud database (Google Sheets) for academic research and trend analysis.
• Session History: Supports viewing history logs within the current browser session.

🛠 Technical Stack

• Frontend Interface: Gradio
• Data Processing: Pandas, Numpy
• Machine Learning: LightGBM (LGBMClassifier)
• Backend Storage: Google Sheets API (via gspread)
• Deployment: Hugging Face Spaces

📁 File Structure

• app.py: The main Gradio application logic.
• DASS_model.bin: The pre-trained machine learning model.
• AutoPreprocess.py: Custom data preprocessing class.
• requirements.txt: List of Python dependencies required for execution.

🔐 Privacy & Ethics

Privacy is our top priority:

1. Anonymity: The "Nickname" input is used solely for the user interface and is NEVER transmitted to the database or used in model calculations.
2. Data Collection: Collected data includes: timestamp, basic demographics (Gender/Age/Family Size), dimensional scores, and the 12 raw responses.
3. Disclaimer: This tool is not a medical device. Results are for reference only and do not constitute a formal diagnosis.

👥 Development Team

• Group 4: Mental Health Risk Prediction Team 🌿
• Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference