metadata
language:
- ja
dataset_info:
features:
- name: document_id
dtype: string
- name: document_name
dtype: string
- name: doc_text
dtype: string
- name: edi_id
dtype: string
- name: security_code
dtype: string
- name: category33
dtype: string
- name: category17
dtype: string
- name: sentence_id
dtype: int64
- name: sentence
dtype: string
- name: opinions
list:
- name: from
dtype: int64
- name: label
dtype: string
- name: polarity
dtype: string
- name: text
dtype: string
- name: to
dtype: int64
splits:
- name: train
num_bytes: 1317601.6
num_examples: 2572
- name: test
num_bytes: 329400.4
num_examples: 643
download_size: 538328
dataset_size: 1647002
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*