# Learning Outcome OS AI-Ready Expanded Dataset v2 Expanded synthetic, CBSE/NCERT LO-aligned, model-ready dataset for LO tagging, Bloom classification, mastery prediction, recommendation, risk prediction, subjective answer scoring, teacher feedback, and digital twin modelling. All names are synthetic. No real PII is included. CSVs are pre-cleaned with stable IDs, target labels, and train/validation/test split columns. ## Tables ### schools.csv - Rows: 3 - Columns: school_id, school_name, board, state, city, region_type, academic_year, is_synthetic ### classes.csv - Rows: 18 - Columns: class_id, school_id, grade, section, class_name, academic_year, class_teacher_id ### subjects.csv - Rows: 9 - Columns: subject_id, grade, subject, subject_code ### chapters.csv - Rows: 52 - Columns: chapter_id, grade, subject, chapter, chapter_order, lo_count, is_active ### teachers.csv - Rows: 54 - Columns: teacher_id, school_id, class_id, teacher_name, grade, section, subject, email, experience_years, ai_feedback_participation_rate ### student_profiles.csv - Rows: 540 - Columns: student_id, school_id, class_id, grade, section, roll_number, student_name, learning_style, learner_archetype, baseline_level, attendance_percentage, assignment_completion_rate, average_login_per_week, inactive_days_last_14, parent_contact_available, is_synthetic, train_split ### learning_outcomes.csv - Rows: 194 - Columns: lo_id, grade, subject, chapter, competency, title, description, difficulty, bloom_level, source_framework, source_pdf, source_pages, alignment_confidence, difficulty_score, bloom_score, embedding_text, is_active, train_split ### lo_dependencies.csv - Rows: 174 - Columns: lo_id, prerequisite_lo_id, relationship_type, strength ### questions.csv - Rows: 5,820 - Columns: question_id, lo_id, grade, subject, chapter, question_text, question_type, difficulty, difficulty_score, bloom_level, bloom_score, correct_answer, rubric, max_marks, source, source_lo_pdf, alignment_confidence, embedding_text, train_split ### question_options.csv - Rows: 7,760 - Columns: question_id, option_label, option_text, is_correct ### content_catalog.csv - Rows: 1,552 - Columns: content_id, lo_id, grade, subject, chapter, title, content_type, target_use, difficulty, duration_minutes, language, description, estimated_mastery_gain, embedding_text, is_active, train_split ### assessments.csv - Rows: 216 - Columns: assessment_id, school_id, class_id, grade, section, subject, assessment_name, assessment_type, scheduled_date, max_marks, question_count, academic_year, train_split ### assessment_questions.csv - Rows: 2,592 - Columns: assessment_id, question_id, question_order ### student_attempts.csv - Rows: 77,760 - Columns: attempt_id, assessment_id, school_id, class_id, student_id, question_id, lo_id, grade, section, subject, question_type, difficulty_score, bloom_score, is_correct, marks_obtained, max_marks, marks_ratio, time_taken_seconds, hint_used, attempt_number, submitted_at, train_split ### initial_mastery_profiles.csv - Rows: 34,920 - Columns: student_id, lo_id, grade, section, subject, chapter, train_split, attempt_count, accuracy, average_marks_ratio, average_time_seconds, hint_usage_rate, mastery_score, status, confidence, last_updated ### mastery_profiles.csv - Rows: 34,920 - Columns: student_id, lo_id, grade, section, subject, chapter, train_split, attempt_count, accuracy, average_marks_ratio, average_time_seconds, hint_usage_rate, mastery_score, status, confidence, last_updated ### engagement_logs.csv - Rows: 108,000 - Columns: engagement_id, student_id, school_id, class_id, activity_date, login_count, active_minutes, content_completed_count, quiz_attempt_count, recommendation_click_count, video_watch_ratio, discussion_posts, device_type, train_split ### risk_profiles.csv - Rows: 1,620 - Columns: risk_prediction_id, student_id, school_id, class_id, grade, section, subject, risk_score, risk_level, risk_label, primary_reasons, recommended_intervention, model_version, generated_at, confidence, train_split ### recommendations.csv - Rows: 8,100 - Columns: recommendation_id, student_id, school_id, class_id, lo_id, content_id, grade, section, subject, recommendation_type, priority, reason, ai_confidence, generated_at, shown_to_student, clicked, is_completed, observed_mastery_gain, model_version, train_split ### teacher_interventions.csv - Rows: 162 - Columns: intervention_id, school_id, class_id, teacher_id, grade, section, subject, lo_id, intervention_type, affected_students, avg_mastery_before, suggested_action, scheduled_week, status, expected_mastery_gain, generated_by_ai, train_split ### subjective_answers.csv - Rows: 12,000 - Columns: answer_id, attempt_id, student_id, question_id, lo_id, grade, subject, question_type, student_answer, model_answer, rubric, max_marks, teacher_marks, ai_predicted_marks, absolute_error, rubric_match_score, concept_coverage_score, feedback_text, teacher_review_required, train_split ### teacher_feedback.csv - Rows: 5,986 - Columns: feedback_id, teacher_id, student_id, related_entity_type, related_entity_id, feedback_type, teacher_rating, correction_required, correction_text, created_at, train_split ### student_digital_twins.csv - Rows: 540 - Columns: digital_twin_id, student_id, school_id, class_id, grade, section, overall_mastery_score, strongest_subject, weakest_subject, top_weak_lo_ids, learning_speed_score, consistency_score, preferred_content_type, current_risk_level, current_risk_score, recommended_next_action, last_updated, train_split ### ai_prediction_logs.csv - Rows: 16,620 - Columns: prediction_log_id, model_name, model_version, entity_type, entity_id, prediction_output, confidence, latency_ms, created_at, train_split ### ml_features_student_subject.csv - Rows: 1,620 - Columns: feature_row_id, student_id, school_id, class_id, grade, section, subject, avg_mastery_score, weak_lo_count, developing_lo_count, mastered_lo_count, avg_confidence, avg_accuracy, avg_marks_ratio, avg_time_seconds, hint_usage_rate, total_attempts, attendance_percentage, assignment_completion_rate, average_login_per_week, inactive_days_last_14, avg_active_minutes, total_logins, avg_video_watch_ratio, total_content_completed, total_quiz_attempts, risk_score, risk_level, risk_label, train_split ### ml_features_student_lo.csv - Rows: 34,920 - Columns: feature_row_id, student_id, lo_id, grade, section, subject, chapter, train_split, attempt_count, accuracy, average_marks_ratio, average_time_seconds, hint_usage_rate, mastery_score, status, confidence, last_updated, school_id, class_id, attendance_percentage, assignment_completion_rate, average_login_per_week, inactive_days_last_14, mastery_label ### training_lo_tagging.csv - Rows: 5,820 - Columns: question_id, question_text, embedding_text, lo_id, grade, subject, chapter, difficulty, difficulty_score, bloom_level, bloom_score, train_split ### training_bloom_classification.csv - Rows: 5,820 - Columns: question_id, question_text, embedding_text, bloom_level, bloom_score, grade, subject, question_type, train_split ### training_risk_prediction.csv - Rows: 1,620 - Columns: feature_row_id, student_id, school_id, class_id, grade, section, subject, avg_mastery_score, weak_lo_count, developing_lo_count, mastered_lo_count, avg_confidence, avg_accuracy, avg_marks_ratio, avg_time_seconds, hint_usage_rate, total_attempts, attendance_percentage, assignment_completion_rate, average_login_per_week, inactive_days_last_14, avg_active_minutes, total_logins, avg_video_watch_ratio, total_content_completed, total_quiz_attempts, risk_score, risk_level, risk_label, train_split ### training_mastery_prediction.csv - Rows: 34,920 - Columns: feature_row_id, student_id, lo_id, grade, section, subject, chapter, train_split, attempt_count, accuracy, average_marks_ratio, average_time_seconds, hint_usage_rate, mastery_score, status, confidence, last_updated, school_id, class_id, attendance_percentage, assignment_completion_rate, average_login_per_week, inactive_days_last_14, mastery_label ### training_answer_scoring.csv - Rows: 12,000 - Columns: answer_id, question_id, student_id, lo_id, grade, subject, question_type, student_answer, model_answer, rubric, max_marks, teacher_marks, ai_predicted_marks, rubric_match_score, concept_coverage_score, teacher_review_required, train_split ### training_recommendation_outcomes.csv - Rows: 8,100 - Columns: recommendation_id, student_id, lo_id, content_id, grade, subject, recommendation_type, priority, ai_confidence, clicked, is_completed, observed_mastery_gain, train_split