An Intelligent Voice-Driven Automated Timetable Generation Framework Using Natural Language Processing and Constraint Optimization
Dr. CH V Phani Krishna,
M Kavya,
M Sabitha Reddy,
P Varun,
M Shiva
Manual timetable generation in educational institutions is labor-intensive, error-prone, and timeconsuming,
often requiring multiple iterations to satisfy complex constraints like teacher availability,
room capacity, and curriculum requirements. This paper proposes an intelligent voice-driven automated
timetable generation framework that integrates Natural Language Processing (NLP) for intuitive voice/
text input and constraint optimization for conflict-free scheduling. Users provide requirements via
natural language voice commands (e.g., "Schedule Mathematics for Class 10 on Monday mornings,
avoid overlapping with Physics lab"). Speech-to-text (STT) converts audio to text, NLP (BERT-based
intent extraction and entity recognition) parses constraints into structured parameters, and a hybrid
solver (Constraint Satisfaction Problem with Genetic Algorithm fallback) optimizes the timetable. The
system generates feasible schedules, handles soft/hard constraints, and provides conflict resolution.
