Upload README.md

README.md

@@ -25,43 +25,69 @@ pretty_name: Digital Habits and Mental Health
 size_categories:
 - 1K<n<10K
 ---
+
 # 🌐 Digital Habits and Mental Health  
 ### Behavioral and Digital Wellbeing Dataset (2025)
 
-A curated dataset exploring how **digital lifestyles** shape **mental wellbeing** — linking screen time, phone use, sleep, and psychological factors such as stress, focus, and happiness.  
-Includes **3,500 anonymized participants** and **24 research-inspired features**, designed for **behavioral research**, **machine learning**, and **explainable AI**.
+A synthetic dataset exploring how **digital lifestyles** shape **mental wellbeing** — linking screen time, phone use, sleep patterns, and psychological factors such as stress, focus, and happiness.  
+Includes **3,500 fully synthetic records** and **24 research-inspired features**, designed for **behavioral analytics**, **machine learning**, and **explainable AI**.
+
+---
+
+## 🔎 Important Note on Scoring
+
+Psychological and behavioral indicators  
+(e.g., `anxiety_score`, `depression_score`, `stress_level`, `happiness_score`, `focus_score`, `productivity_score`, `digital_dependence_score`)  
+are generated as **continuous synthetic scores modeled on a broad 0–100 range**, not fixed 0–10 Likert items.
+
+This provides richer variance and enhances suitability for ML models and interpretability techniques.
 
 ---
 
 ## 📘 Dataset Overview
 
 | Field | Description |
-|:--|:--|
+|------|-------------|
 | **File name** | `Data.csv` |
 | **Rows** | 3,500 |
 | **Columns** | 24 |
 | **Target** | `high_risk_flag` |
+| **Type** | Tabular (Synthetic) |
 
 ---
 
 ## 🧠 Feature Groups
 
-**Demographics:** age · gender · region · income_level · education_level  
-**Digital Behavior:** daily_screen_time · phone_unlocks · notifications_per_day · social_media_hours · study_time · work_hours  
-**Mental Health Indicators:** stress_level · anxiety_level · depression_level · happiness_score · focus_score · productivity_score  
-**Derived Ratios:** screen_to_sleep_ratio · social_work_ratio · wellbeing_index  
-**Target:** `high_risk_flag` — binary wellbeing-risk indicator (0 = low, 1 = high)
+### Demographics  
+`age`, `gender`, `region`, `income_level`, `education_level`
+
+### Digital Behavior  
+`device_hours_per_day`, `phone_unlocks`, `notifications_per_day`,  
+`social_media_hours`, `daily_screen_time`, `study_time`, `work_hours_per_day`
+
+### Mental Health Indicators  
+`stress_level`, `anxiety_score`, `depression_score`,  
+`happiness_score`, `focus_score`, `productivity_score`, `sleep_quality`
+
+### Additional Behavioral Metrics  
+`digital_dependence_score`, `risk_exposure_score`
+
+### Target  
+`high_risk_flag` — binary wellbeing-risk indicator (0 = low, 1 = high)
 
 ---
 
 ## 🎯 Target Definition
 
-The target variable **`high_risk_flag`** labels individuals at elevated mental-health risk, defined by a composite of:
+The target variable **`high_risk_flag`** identifies individuals with elevated wellbeing risk.  
+It is computed through a composite scoring process blending:
+
 - High digital engagement (screen time, unlocks, notifications)  
-- Elevated stress or anxiety  
-- Lower happiness/focus scores  
+- Elevated stress/anxiety  
+- Low focus or happiness  
+- Behavioral intensity patterns  
 
-Distribution: roughly **15–20 % high-risk**, aligning with behavioral research estimates.
+Distribution: **15–20% high-risk**, aligned with behavioral research estimates.
 
 ---
 
@@ -71,34 +97,43 @@ Distribution: roughly **15–20 % high-risk**, aligning with behavioral research
 from datasets import load_dataset
 
 dataset = load_dataset("TarekMasryo/digital-habits-mental-health")
+df = dataset["train"].to_pandas()
+```
 
 ---
 
 ## 🔬 Research & Applications
 
-- Predict wellbeing risk from digital patterns  
+- Predict digital wellbeing risk  
 - Correlate stress, sleep, and screen exposure  
-- Build explainable models (SHAP / LIME)  
-- Segment behavioral profiles by lifestyle balance  
-- Apply calibration or threshold tuning for decision systems  
+- Build explainable AI models (SHAP / LIME)  
+- Behavioral segmentation and pattern analysis  
+- Threshold tuning and calibration for decision systems  
 
 ---
 
 ## 🧩 Reproducibility
 
 - No missing or duplicate values  
-- Deterministic preprocessing and schema  
+- Deterministic synthetic generation  
+- Fully ML-ready schema  
 - Compatible with Kaggle, Colab, and Jupyter  
 
 ---
 
----
-
 ## 🧭 Ethical Considerations
-Educational/research dataset only — **not** for clinical use, diagnosis, or treatment.
+
+This dataset is **synthetic** and intended for **educational and research purposes only** —  
+not for clinical, diagnostic, or therapeutic use.
+
+---
 
 ## 📚 Citation
+
 Please cite the dataset URL on Hugging Face and the license below when using this data.
 
-## 📜 License
-**CC BY 4.0 (Attribution)** — Free to share and adapt with attribution.
+---
+
+## 📜 License  
+**CC BY 4.0 (Attribution Required)**  
+Free to use, share, and modify with proper attribution.