Dark Mode
SnappFood Persian Sentiment Reviews
Food & Beverage Consumption
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
A valuable collection of cleaned, Persian-language user reviews focused on the sentiment expressed towards a food delivery service. The original data contained approximately 70,000 samples, which were processed down to 65,973 high-quality records. This resource is essential for developing robust sentiment analysis models in Persian, as significant preprocessing steps were undertaken, including text normalization, removal of noise elements (links, punctuation, stopwords), and outlier removal via BoxPlot analysis. These cleaning methods were employed specifically to boost the learning capability of machine learning models.
Columns
- comment: The raw, original text of the user review.
- label: The numerical sentiment category for the review, where 0 indicates Negative (originally "Sad") sentiment and 1 indicates Positive (originally "Happy") sentiment.
- comment_length: The measured length of the original user comment text.
- comment_cleaned: The preprocessed version of the comment, suitable for direct input into models.
Distribution
This dataset contains 65,973 samples of user feedback. The data is highly balanced across the two sentiment labels: 33,675 records are classified as Positive (1), and 32,298 records are classified as Negative (0). The statistics for the label column show a Mean of 0.51 and a standard deviation of 0.5. The lengths of the original comments vary widely, ranging from a minimum of 11 characters to a maximum of 226 characters, with an average length of 76.1 characters. The data is available as a CSV file (
cleaned_snappfood.csv).Usage
Ideal applications for this dataset include training Natural Language Processing (NLP) models for sentiment classification specific to the Persian language and food delivery context. It can be used for monitoring customer satisfaction trends, performing detailed linguistic analysis of user feedback in Farsi, and benchmarking new text classification algorithms.
Coverage
The data scope is focused on user feedback collected from the SnappFood food delivery service. The language is exclusively Persian (Farsi). The dataset is static and has an expected update frequency of 'Never'. The content deals specifically with consumer experiences related to ordering food, encompassing quality, service, and delivery comments.
License
CC0: Public Domain
Who Can Use It
- Data Scientists and Machine Learning Engineers: For building, training, and fine-tuning sentiment analysis models tailored for low-resource or domain-specific languages like Persian.
- NLP Researchers: For conducting linguistic studies on sentiment expression, slang, and informal language use within user-generated content in Persian.
- Market Analysts: To gain insights into customer satisfaction drivers and pain points within the food delivery sector.
Dataset Name Suggestions
- SnappFood Persian Sentiment Reviews
- Cleaned Farsi Food Delivery Sentiment Corpus
- Persian NLP Customer Feedback Dataset
Attributes
Original Data Source: SnappFood Persian Sentiment Reviews
QUALITYLoading...