metadata

license: mit
dataset_info:
  features:
    - name: normalized_review_text
      dtype: string
    - name: rating
      dtype: string
  splits:
    - name: train
      num_bytes: 19637178
      num_examples: 352151
  download_size: 10181120
  dataset_size: 19637178
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
task_categories:
  - text-classification
  - token-classification
language:
  - uz
tags:
  - sentiment_analysis
pretty_name: Sentiment analysis for Uzbek language.
size_categories:
  - 100K<n<1M

Uzum Market Sentiment Analysis Dataset

About the Dataset

This dataset is created for performing sentiment analysis on comments from Uzum Market. The dataset allows for evaluating the sentiments in the comments and categorizing them into various ratings.

Data Structure

The data is provided in JSON format with the following structure:

{
    "features": ["normalized_review_text", "rating"],
    "num_rows": 352151
}

Ratings are defined as follows:

rating_to_label = {
    1: 'very poor',
    2: 'poor',
    3: 'fair',
    4: 'good',
    5: 'excellent'
}

Preparation Guidelines

The following guidelines were followed in preparing this dataset:

Normalized review texts and their ratings were extracted.
Each rating is associated with different quality indicators (good, average, bad, etc.).

Examples

Here are some examples from the dataset:

normalized_review_text: ["The product is very good, I recommend it!", "The service quality is bad, not satisfied."]
rating: [5, 1]

Download

You can download this dataset using the following code:

from datasets import load_dataset

dataset = load_dataset("risqaliyevds/sentiment-analysis-uzbek")

License

This dataset is provided as open source and is available for free use by all users.

Contact

If you have any questions or need more information, please contact us. LinkedIn: Riskaliev Murad