bonmubyte/kms_broadcast_eval

## Dataset: KMS Broadcast Eval

### Description

This dataset contains financial broadcast evaluation data for compliance checks. The dataset includes labeled data to identify violations based on specific criteria.

### Files

- **kms_broadcast_eval.csv**: The main dataset containing context, violation status, reasons, and categories.

### Label Definitions

- **0**: Not violated
- **1**: Violated

### Columns

- **id**: Unique identifier for each row
- **context**: The actual advertisement text that is being evaluated
- **violation_status**: Indicates whether there is a violation in the given context (0 for not violated, 1 for violated)
- **violation_reason**: The reasons for the violation, listing the specific phrases that are problematic
- **category_1**: Violation status for "단정적인 표현" or "절대적 수익" (0 for not violated, 1 for violated)
- **category_2**: Violation status for "손실 보전" or "수익 보장" (0 for not violated, 1 for violated)
- **category_3**: Violation status for "근거없거나 과도한 최상급 혹은 긍정적 표현" (0 for not violated, 1 for violated)
- **category_4**: Violation status for "수익률 관련 금지표현 및 권장표현" (0 for not violated, 1 for violated)
- **category_5**: Violation status for "금융 상품 간의 비교 표현" (0 for not violated, 1 for violated)

原始信息汇总

数据集: KMS Broadcast Eval

描述

该数据集包含用于合规检查的金融广播评估数据。数据集包括标记数据，用于根据特定标准识别违规行为。

文件

• kms_broadcast_eval.csv: 主要数据集，包含上下文、违规状态、原因和类别。

标签定义

• 0: 未违规
• 1: 违规

列

• id: 每行的唯一标识符
• context: 正在评估的实际广告文本
• violation_status: 指示给定上下文中是否存在违规（0表示未违规，1表示违规）
• violation_reason: 违规原因，列出具体的问题短语
• category_1: "단정적인 표현" 或 "절대적 수익" 的违规状态（0表示未违规，1表示违规）
• category_2: "손실 보전" 或 "수익 보장" 的违规状态（0表示未违规，1表示违规）
• category_3: "근거없거나 과도한 최상급 혹은 긍정적 표현" 的违规状态（0表示未违规，1表示违规）
• category_4: "수익률 관련 금지표현 및 권장표현" 的违规状态（0表示未违规，1表示违规）
• category_5: "금융 상품 간의 비교 표현" 的违规状态（0表示未违规，1表示违规）