游戏类京东直播带货违法监测预警数据

对采集的京东平台带货品类为游戏类的达人直播视频内容进行转译分析，对达人口播语言内容进行处理、分析，根据达人在直播过程中对预先设置的违规敏感词（比如：玩到停不下来、不限时、一夜暴富、学生、真金白银、免费兑换、18禁、充的越多越厉害、百万收入不是梦、畅玩等）违反的次数和频率，依据触发条件规则提出警告或处理。为服务辖区市场监督局管理区域内规范企业京东直播行为，提供数据支持。将采集完成的直播视频进行进行预处理，第一步：基于原始视频文件，以最大10分钟单位对原始视频进行切片。第二步：对于已完成的切片视频，进行视频内容转语音操作。第三步：对于已完成视频转语音操作的切片，进行语音转文本操作。第四步：使用OCR技术对原始视频中抓取的图片进行文字提取操作。第五步：将所得到的文字内容与违法预警关键词库进行匹配。最终运用多标准决策分析模型，对主播在直播过程中出现的违规语句进行分析计算，得出违法预警值和是否预警判断。
违法预警值=（违法预警单关键词命中次数*0.25）+（违法预警组合关键词命中次数* 0.3）+（图片识别命中预警组合关键词个数*0.35）+（直播间近一个月历史违规记录数*0.1）
通过公式计算出最终违法预警值，违法预警值 ≤1 时，不触发预警提示，违法预警值 ＞1 时触发违法预警提示。

This study performs transcription and analysis on collected game-category sponsored livestream videos of JD.com platform streamers, and processes and analyzes the oral language content of the streamers. Warnings or corresponding corrective actions are issued in accordance with trigger condition rules, based on the frequency and count of violations of pre-configured sensitive violation keywords (e.g., can't stop playing, no time limit, get rich overnight, students, real cash, free redemption, 18+, the more you top up the stronger, millions in income is no dream, unrestricted play, etc.) during the livestream. This dataset provides data support for market supervision bureaus under their jurisdiction to standardize the livestreaming behaviors of enterprises operating on JD.com.

The collected livestream videos undergo the following preprocessing workflow:
1. Segment the original video files into clips with a maximum duration of 10 minutes per clip.
2. Extract audio from the clipped videos (i.e., convert video content to speech).
3. Conduct speech-to-text conversion on the extracted audio from the clipped videos.
4. Perform text extraction from images captured from the original videos using Optical Character Recognition (OCR) technology.
5. Match all obtained text content against the pre-built violation warning keyword database.

Finally, a multi-criteria decision analysis (MCDA) model is employed to analyze and calculate the violating statements made by the streamer during the livestream, to derive the final violation warning score and determine whether to trigger a warning prompt.

The violation warning score is calculated using the following formula:
Violation Warning Score = (Number of hits for individual violation warning keywords × 0.25) + (Number of hits for combined violation warning keywords × 0.3) + (Number of combined warning keywords identified via image recognition × 0.35) + (Number of historical violation records of the livestream room in the past 30 days × 0.1)

The final violation warning score is calculated via the above formula. No warning prompt is triggered when the score is ≤ 1, while a violation warning prompt is activated when the score exceeds 1.

搜集汇总

数据集介绍

特点

该数据集主要用于监测京东平台游戏类直播中的违法行为，包含1005条记录，每季度更新。通过分析直播内容中的违规敏感词，结合算法规则计算违法预警值，为市场监督局提供数据支持。