Siki-77/yelp_3classes

该数据集包含来自Yelp的评论数据，每条评论都有一个标签表示情感倾向（负面、中性、正面）。数据集分为训练集和测试集，训练集包含650,000条评论，测试集包含50,000条评论。标签0表示负面（来自0和1星评价），1表示中性（来自2星评价），2表示正面（来自3和4星评价）。

This dataset contains review data from Yelp, with each review labeled for sentiment (negative, neutral, positive). The dataset is divided into a training set with 650,000 reviews and a test set with 50,000 reviews. Label 0 indicates negative (from 0 and 1-star ratings), 1 indicates neutral (from 2-star ratings), and 2 indicates positive (from 3 and 4-star ratings).

原始信息汇总

数据集概述

数据集配置

• config_name: default

  • features:
    • label: int64
    • t_label: string
    • text: string
  • splits:
    • train:
      • num_bytes: 491481554
      • num_examples: 650000
    • test:
      • num_bytes: 37861188
      • num_examples: 50000
  • download_size: 323230667
  • dataset_size: 529342742

• config_name: test

  • features:
    • label: int64
    • t_label: string
    • text: string
  • splits:
    • train:
      • num_bytes: 37861188
      • num_examples: 50000
  • download_size: 23535516
  • dataset_size: 37861188

数据文件路径

• config_name: default
  • data_files:
    • split: train
      • path: data/train-*
    • split: test
      • path: data/test-*

标签定义

• 0: Negative (来自星级 0 和 1)
• 1: Neutral (来自星级 2)
• 2: Positive (来自星级 3 和 4)

示例

json { "label": 0, "t_label": "Negative", "text": "dr. goldberg offers everything i look for in a general practitioner. hes nice and easy to talk to without being patronizing; hes always on time in seeing his patients; hes affiliated with a top-notch hospital (nyu) which my parents have explained to me is very important in case something happens and you need surgery; and you can get referrals to see specialists without having to see him first. really, what more do you need? im sitting here trying to think of any complaints i have about him, but im really drawing a blank." }

数据来源

• 来源: Yelp
• 引用: Xiang Zhang, Junbo Zhao, Yann LeCun. Character-level Convolutional Networks for Text Classification. Advances in Neural Information Processing Systems 28 (NIPS 2015).