PID 78

SORTIFY AI-Powered Sorting Hat Personality-Based House Prediction

A revolutionary AI-driven system that combines personality quizzes, emotional analysis, physiological data, and cognitive testing to accurately determine your Hogwarts house and provide personalized career insights.

Explore Project View Interfaces

Innovators

Shahidatul Hidayah binti Ahmad Faizal Undergraduate

Assoc. Prof. Dr. Azizi bin Ab Aziz Supervisor

Scroll to explore

Background

The Problem

Traditional personality assessments face significant limitations that impact their reliability and effectiveness.

Self-Reported Bias: Users may provide answers they think are expected rather than honest responses
Static Assessment: Cannot capture real-time emotional states or physiological responses
Limited Context: Fails to account for cognitive abilities and emotional intelligence
Subjective Results: Lacks objective data validation from multiple sources

Our Solution

Sortify revolutionizes personality assessment by integrating multiple data sources for comprehensive analysis.

Multimodal Analysis: Combines personality quizzes with AI-driven emotional and physiological analysis
Real-Time Data: Captures live emotional states through facial expression analysis and HRV monitoring
Cognitive Integration: Incorporates IQ testing and cognitive assessments for holistic evaluation
Career Relevance: Provides actionable insights aligned with personalized career pathways

Methodology

Process Iteration Model

Our development follows a systematic Design Science Research Methodology (DSRM) approach, ensuring rigorous validation and continuous improvement.

System Algorithm

The complete workflow from initial personality assessment to final house assignment and personalized report generation.

Fusion Model: Hogwarts House Assignment

h_x^new(t) = h_x^old(t) + Σ f_x(Fr, Em, Cn, Cl, Iq)

h_x^new(t): Updated house score for house x at time t

h_x^old(t): Baseline house score from personality quiz

f_x: Feature function combining:

Fr = Fear response
Em = Empathy level
Cn = Cunning assessment
Cl = Cleverness score
Iq = IQ test results

Evaluation

Data Analytics

Comprehensive model comparison showing the superior performance of our weighted voting ensemble approach.

Top 3 Ensemble Models

Random Forest 81.00%
Extra Trees 81.00%
XGBoost 80.67%

Weighted Voting Ensemble: 81.33% Accuracy

User Testing

User experience evaluation using AttrakDiff methodology measuring pragmatic and hedonic quality.

Scale	Mean	Std. Dev.	Rating
Pragmatic Quality	2.233	0.723	Good
Hedonic Quality	2.500	0.582	Excellent
Overall	2.367	0.598	Good

Key Finding: Users rated the system highly for both usability (pragmatic) and enjoyment (hedonic) aspects, indicating successful balance between functionality and engagement.

Results

User testing results demonstrate the effectiveness of our AI-powered sorting system. The distribution shows clear house preferences among participants.

House Distribution

53.3%

Gryffindor

40%

Hufflepuff

6.7%

Ravenclaw

Key Insights

Gryffindor Dominance

Over half of participants (53.3%) were sorted into Gryffindor, indicating strong alignment with bravery and courage traits.

Hufflepuff Popularity

40% of users identified with Hufflepuff values, showing appreciation for loyalty and fairness.

Ravenclaw Representation

6.7% were sorted into Ravenclaw, reflecting the specialized nature of wisdom and learning traits.

Interfaces

Our user-friendly interface design ensures an engaging and intuitive experience throughout the assessment process. Each screen is carefully crafted to guide users seamlessly.

Homepage

Personalized greeting
Quick access to quiz
Previous results display
Career pathway links

Stimulus Page

Video stimulus playback
Live emotion detection
HRV monitoring
Interactive quizzes

Results Page

House prediction display
Probability breakdown
Personalized report
Career recommendations

Video Demo

Watch our comprehensive demonstration of the Sortify system in action, showcasing the complete assessment process from personality quiz to final house assignment.

Personality Assessment

Complete Big Five Inventory questionnaire

Stimulus Evaluation

Real-time emotion and HRV monitoring

IQ Testing

Cognitive ability assessment

House Assignment

AI-powered house prediction and report

Impact & Future

Impact

Improved Accuracy

Integrates cognitive, emotional, and physiological data for more reliable personality classification (81.33% accuracy).

Career Guidance

Provides meaningful real-world guidance with personalized career and self-development insights.

User Engagement

High hedonic quality rating (2.500) demonstrates excellent user satisfaction and engagement.

Novelty

Multimodal AI

First system to combine personality quizzes, HRV analysis, real-time facial expression recognition, and IQ testing in a single framework.

Real-Time Processing

Captures and processes emotional and physiological data in real-time during stimulus evaluation.

Ensemble Approach

Weighted voting ensemble of top-performing models ensures robust and accurate predictions.

Future Works

Dataset Expansion

Expand training dataset to improve model robustness and generalizability across diverse populations.

Mobile Application

Develop native mobile app for enhanced accessibility, real-time sensing, and improved user experience.

Model Refinement

Continuous improvement of ML models through advanced techniques and larger datasets.

Commercializability

Educational Applications

Suitable for fun educational apps, helping students understand personality traits and career alignment.

HR Applications

Valuable tool for Human Resources departments in talent assessment and team building.

Market Ready

Proven accuracy and user satisfaction make it ready for commercial deployment.