AVRS-vocab

Dataset of English vocabulary for an AI Vocabulary Recommendation System (AVRS). It contains words with topic labels, difficulty levels, word length, and syllable counts for downstream tasks such as KNN-based recommendation and difficulty classification.

• Dataset page: https://huggingface.co/datasets/klcharizardqwq/AVRS-vocab

Files

• vocab.csv — main vocabulary table used by the notebooks

Schema

Columns may appear in title or lowercase forms depending on the source file. Typical columns:

• Word or word: the vocabulary token (string)
• Topic or topic: semantic topic/category (e.g., School, Travel, Daily, Sport, Health, Technology, ...)
• Difficulty or difficulty: one of EASY, MEDIUM, HARD
• Word length or Word Length or word length: character length of the word (int)
• Number of Syllables or no. of syllables: estimated syllable count (int)

Example row (CSV):

word,Topic,difficulty,word length,no. of syllables
abacus,technology,MEDIUM,6,3

Intended uses

• Topic-aware vocabulary recommendation (e.g., via KNN/embeddings)
• Difficulty classification (e.g., DecisionTreeClassifier)
• Educational analytics and visualization

Loading examples

Python (pandas) using huggingface_hub:

from huggingface_hub import hf_hub_download
import pandas as pd

path = hf_hub_download(repo_id="klcharizardqwq/AVRS-vocab", filename="vocab.csv", repo_type="dataset")
df = pd.read_csv(path)
print(df.head())

Python (datasets):

from datasets import load_dataset

ds = load_dataset("klcharizardqwq/AVRS-vocab", data_files={"train": "vocab.csv"})
print(ds["train"][0])

Notes

• Topics in vocab.csv are curated for classroom-friendly categories (e.g., technology, science, nature, people).
• Syllable counts use pyphen when available, with a robust heuristic fallback.

License

Please refer to the dataset repository’s license metadata on Hugging Face (currently listed as Apache-2.0 on the dataset page) and comply with those terms.

If you need a different license for this dataset, update the repo’s license setting and the README accordingly.