README.md

metadata

annotations_creators:
  - no-annotation
language:
  - vi
language_creators:
  - found
license: cc-by-nc-4.0
multilinguality:
  - monolingual
size_categories:
  - n<1K
source_datasets: []
task_categories:
  - multiple-choice
task_ids:
  - multiple-choice-qa

AIO-2025 Vietnamese ML/DS Dataset

Vietnamese multiple-choice question dataset for AI, Machine Learning, and Data Science education.

Quick Start

Python / HuggingFace

from datasets import load_dataset

ds = load_dataset("vudang449/AIO-2025-Vietnamese-ML-DS")
print(ds["train"][0]["question"])

Local Files

import json

with open("aio2025_final_clean.jsonl") as f:
    for line in f:
        record = json.loads(line)

Dataset Summary

Metric	Value
Total records	197
Balanced answers	A: 50, B: 50, C: 50, D: 47
Pass rate	100% (quality-verified)
Language	Vietnamese
License	CC BY-NC 4.0

Fields

Field	Type	Description
id	string	Unique record ID
question	string	Question text
option_a	string	Choice A
option_b	string	Choice B
option_c	string	Choice C
option_d	string	Choice D
answer	string	Correct answer: A / B / C / D
answer_index	int	Numeric index (0=A, 1=B, 2=C, 3=D)
source	string	quiz / exam / vietnamese_exam
module	string	AIO course module (Module_3A–Module_14, Module_3B)
difficulty	string	EASY / MEDIUM / HARD

Distribution

By Answer (balanced)

• A: 50 (25.4%), B: 50 (25.4%), C: 50 (25.4%), D: 47 (23.9%)

By Difficulty

• MEDIUM: 118 (59.9%)
• EASY: 71 (36.0%)
• HARD: 8 (4.1%)

By Module

• Module_3A: 2, Module_3B: 5, Module_4: 17, Module_5: 26
• Module_6: 24, Module_7: 27, Module_8: 10, Module_9: 27
• Module_10: 13, Module_11: 5, Module_12: 4, Module_13: 2
• Module_14: 6, Quizzes: 29

Quality Assurance

Records removed during cleaning (problematic records):

• [Blank] placeholder markers
• Cross-question dependencies
• Image/diagram dependencies
• Code in options (notebook exercises)
• Math fragment leakage in options
• Option contamination
• Schema inconsistencies

Citation

@dataset{aio2025vietnamese,
  title  = {AIO-2025 Vietnamese ML/DS Dataset},
  author = {AIO Course},
  year   = {2025},
  version = {v2.0.0},
  license = {CC BY-NC 4.0},
}