Indonesian Hospital Review Dataset XLSX [25K]

The dataset used in this study consists of textual reviews and corresponding sentiment labels related to hospital services in Indonesia. The reviews were collected from the Google Reviews platform, covering 150 hospitals located across the main islands of Indonesia (Java, Bali, Sumatera, Kalimantan, and Sulawesi). Data collection was conducted manually using the Instant Data Scraping tool, which allows the selection of hospital profiles and the extraction of specific columns, including the review text and its metadata. Each scraping process for one hospital location took approximately 20 minutes, after which the raw data was cleaned and preprocessed for further analysis.

The review column contains user-generated feedback in the Indonesian language, with lengths ranging from short comments (~10 characters) to more detailed opinions (~300 characters on average), while the sentiment column represents manually assigned sentiment polarity (e.g., positive or negative) based on the overall tone and content of the review. The dataset contains approximately 25,000 entries, with the number of reviews per hospital ranging from 170 to 1,050.