perfume-dataset

Perfume Dataset 是一个从 'parfumo_tidytuesday' 爬取的香水记录数据集，已转换为 Hugging Face 兼容格式。数据集包含两种文件格式：Parquet 和 JSON Lines，均为训练集数据。每条记录对应一个爬取的香水条目，并保留了爬取步骤中的原始字段。数据集规模为 42,095 条训练样本。该数据集适用于香水相关的研究或分析任务，如香水属性分析、推荐系统开发等。

The Perfume Dataset is a perfume record dataset scraped from 'parfumo_tidytuesday' and converted to Hugging Face-compatible format. It includes two file formats: Parquet and JSON Lines, both of which are training data. Each record corresponds to a scraped perfume entry, retaining all original fields from the crawling process. The dataset contains 42,095 training samples, and is applicable to perfume-related research or analysis tasks such as perfume attribute analysis and recommendation system development.