| --- |
| license: mit |
| language: |
| - en |
| metrics: |
| - precision |
| - recall |
| - f1 |
| - accuracy |
| - roc_auc |
| pipeline_tag: tabular-classification |
| tags: |
| - classification |
| - psychology |
| --- |
| |
| # Model Card for Infinitode/BPPM-OPEN-ARC |
|
|
| Repository: https://github.com/Infinitode/OPEN-ARC/ |
|
|
| ## Model Description |
|
|
| OPEN-ARC-BPP is a simple XGBClassifier model developed as part of Infinitode's OPEN-ARC initiative. It was designed to determine a person's basic personality type (introvert/extrovert) based on several personal questions. |
|
|
| **Architecture**: |
|
|
| - **XGBClassifer**: `n_estimators=200`, `learning_rate=0.1`, `max_depth=4`, `eval_metric="logloss"`, `use_label_encoder=False`, `random_state=42`. |
| - **Framework**: XGBoost |
| - **Training Setup**: Trained using selected params. |
|
|
| ## Uses |
|
|
| - Identifying a person's basic personality type through a series of personal questions. |
| - Enhancing knowledge and research in psychology and human behavior. |
|
|
| ## Limitations |
|
|
| - Might lead to inaccurate evaluations of personality types because of exceptions and inconsistencies. |
|
|
| ## Training Data |
|
|
| - Dataset: Personality Dataset (introvert or Extrovert) dataset from Kaggle. |
| - Source URL: https://www.kaggle.com/datasets/hardikchhipa28/personality-dataset-introvert-or-extrovert |
| - Content: Personality-determining features such as hours spent alone per day, etc. |
| - Size: The size is unknown as the dataset is no longer publicly available. |
| - Preprocessing: Preprocessing involved techniques such as number and category imputation, one-hot encoding, and label encoding. |
|
|
| ## Training Procedure |
|
|
| - Metrics: accuracy, precision, recall, F1, ROC AUC |
| - Train/Testing Split: 75% train, 25% testing. |
|
|
| ## Evaluation Results |
|
|
| | Metric | Value | |
| | ------ | ----- | |
| | Testing Accuracy | 92.0% | |
| | Testing Weighted Average Precision | 92% | |
| | Testing Weighted Average Recall | 92% | |
| | Testing Weighted Average F1 | 92% | |
| | Testing ROC AUC | 95.5% | |
|
|
| ## How to Use |
|
|
| ```python |
| import joblib |
| import pandas as pd |
| import numpy as np |
| |
| # Load artifacts |
| art = joblib.load("personality_artifacts.pkl") |
| model = art["model"] |
| num_imputer = art["num_imputer"] |
| cat_imputer = art["cat_imputer"] |
| ohe = art["ohe"] |
| le = art["label_encoder"] |
| num_cols = art["num_cols"] |
| cat_cols = art["cat_cols"] |
| |
| def ask(prompt, cast=str, options=None): |
| """Tiny helper to get clean input.""" |
| while True: |
| try: |
| val = cast(input(prompt)) |
| if options and val not in options: |
| raise ValueError(f"Must be one of {options}") |
| return val |
| except Exception as e: |
| print(f"Error: {e}. Try again.\n") |
| |
| def predict_personality(): |
| print("\n🎭 Introvert vs Extrovert Predictor") |
| print("Answer a few quick questions:\n") |
| |
| # Gather answers |
| answers = { |
| "Time_spent_Alone": ask("Hours spent alone per day (0‑24): ", float), |
| "Stage_fear": ask("Stage fear? (Yes/No): ", str.title, ["Yes", "No"]), |
| "Social_event_attendance":ask("Social events per week (0‑10): ", int), |
| "Going_outside": ask("Trips outside per day (0‑10): ", int), |
| "Drained_after_socializing": ask("Feel drained after socializing? (Yes/No): ", |
| str.title, ["Yes", "No"]), |
| "Friends_circle_size": ask("Number of close friends (0‑30): ", int), |
| "Post_frequency": ask("Social‑media posts per week (0‑30): ", int), |
| } |
| |
| # Build one‑row DataFrame in correct column order |
| row = pd.DataFrame([answers])[num_cols + cat_cols] |
| |
| # --- Re‑run exact preprocessing ------------------------------------------------- |
| X_num = num_imputer.transform(row[num_cols]) |
| X_cat = cat_imputer.transform(row[cat_cols]) |
| X_cat_enc = ohe.transform(X_cat) |
| X_ready = np.hstack([X_num, X_cat_enc]) |
| |
| # --- Predict -------------------------------------------------------------------- |
| proba = model.predict_proba(X_ready)[0] |
| idx = proba.argmax() |
| pred_label = le.inverse_transform([idx])[0] |
| confidence = proba[idx] |
| |
| print(f"\n🔮 You are likely an **{pred_label}** (confidence {confidence:.0%}).") |
| |
| predict_personality() |
| ``` |
|
|
| ## Contact |
|
|
| For questions or issues, open a GitHub issue or reach out at https://infinitode.netlify.app/forms/contact. |
