Movie Streaming Customer Support Messages with Toxicity Labels

Dataset Review Movie is a collection of labeled user messages to an online movie theater's customer support service, intended for training text toxicity classifiers. The dataset was created by Gloomreach and was last updated on July 15, 2026. It includes fields for the message text, a toxicity label, a data source, and a toxicity category.

Use Cases

• Train a binary text classifier to detect toxic vs. non-toxic messages based on the 'label' field.
• Analyze sub-categories of toxic content in customer support based on the 'category' field for toxic examples.
• Benchmark toxicity detection models on a domain-specific dataset from a movie streaming platform.
• Study the characteristics of user complaints and feedback in online entertainment services.

Strengths

• Dataset is explicitly designed for training toxicity classifiers, providing a clear purpose.
• Includes a 'category' field for toxic examples, which suggests finer-grained labeling.

Limitations

• Row count, file formats, and license information are unknown, limiting suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Data sources include API-generated content and open datasets, which may introduce specific biases.

Provenance

Source

Gloomreach, sourced from manual taxonomy, API generation, and open datasets from Kinopoisk.

Collection Method

Likely a combination of manual labeling and automated collection from specified sources.

Freshness

Last updated 2026-07-15 20:27:12; freshness should be verified.