Hotel Reviews in Indonesia with Sentiment Labels

Featuring reviews of hotels in Indonesia, with each entry labeled as positive (1) or negative (0). It includes fields for an unique review_id and the review_text. The dataset size is categorized as between 10,000 and 100,000 rows.

Use Cases

• Train a text classifier to predict the sentiment category (positive/negative) from review_text.
• Analyze the distribution of positive and negative category labels across the dataset.
• Perform topic modeling on review_text to identify common themes in hotel feedback.
• Use review_id to track and deduplicate user reviews in downstream analysis.

Strengths

• Includes explicit sentiment labels (category) for each review_text.
• Dataset size is categorized as 10K<n<100K, indicating a substantial collection.
• Data is focused on a specific geography: Indonesia.

Limitations

• The exact row count and column details are unknown, limiting precise scalability assessment.
• The binary sentiment category may oversimplify more nuanced opinions in the review_text.
• Data is from a single source and may not represent all hotel review patterns in Indonesia.

Provenance

Source

huggingface, author rakkaalhazimi

Freshness

Last updated on 2022-03-12.

Geography

Indonesia