Dataset

restaurant_reviews

The restaurant_reviews dataset is a small text corpus of roughly one thousand customer review snippets paired with binary sentiment labels, published on the Hugging Face Hub under the user account pachequinho. It is distributed in CSV format with an Apache-2.0 license declaration.

Open on Hugging Face

Data preview

Subset

Loading preview...

Dataset card

What this dataset contains

The file holds a single training split of about 1,000 rows, each carrying a free-text review and an associated integer sentiment label. According to the schema, the label column uses int64 values, and the text is stored as a string Value field, making it suitable for lightweight text-classification experiments. Sample entries indicate that positive reviews are tagged with a 1 and negative reviews with a 0, suggesting a binary polarity task.

Structure and fields

Two columns are defined: text, a string field holding the raw review content, and labels, an int64 field carrying the sentiment class. The data is delivered as a single CSV train split with no separate validation or test partition, and no additional metadata columns such as reviewer id, timestamp, or restaurant identifier are present. The dataset is tagged with the Hugging Face datasets, pandas, polars, and mlcroissant libraries, indicating it conforms to common machine-learning tooling conventions.

How analysts use it

Practitioners typically use the pair of text and labels columns to train and benchmark small classifiers for restaurant-level sentiment detection, often as a teaching or prototyping resource rather than a production feed. Because the corpus is tiny and lacks restaurant identifiers or timestamps, it is of limited direct use for investment workflows; it would mainly serve as a sandbox example for analysts exploring natural-language-processing pipelines before applying larger sentiment datasets to consumer or hospitality-sector signals.

Documentation & references 1