酒类快手直播带货违法监测预警数据

对采集的快手平台带货品类为酒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：不宿醉，喝不醉，不上头，醒酒快，温肾，除积冷，消除紧张、消除焦虑，增加体力，改善疲劳感，再喝一杯，千杯不醉等）违反的次数和频率，依据触发条件规则提出警告或处理。为诸暨市市场监督局管理区域内规范企业快手直播行为，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 dataset focuses on the transcription and analysis of collected live stream video content from liquor-category Kuaishou platform live streamers, as well as the processing and analysis of the verbal content uttered by the streamers during their broadcasts. It counts and analyzes the frequency and number of violations against pre-set sensitive violation keywords (including but not limited to: no hangover, cannot get drunk, no head rush, quick sobering up, warm the kidneys, remove accumulated cold, relieve tension and anxiety, boost physical stamina, alleviate fatigue, have another drink, never get drunk even after drinking a thousand cups). Corresponding warnings or penalties will be issued based on the triggering condition rules, providing data support for the Zhuji Municipal Market Supervision Administration to regulate the live streaming behaviors of enterprises within its jurisdiction on the Kuaishou platform.

The collected live stream videos undergo the following preprocessing steps:
Step 1: Slice the original video files into segments with a maximum duration of 10 minutes each.
Step 2: Extract and convert the audio content from each sliced video segment.
Step 3: Perform speech-to-text transcription on the extracted audio files.
Step 4: Extract text from images captured from the original videos using Optical Character Recognition (OCR) technology.
Step 5: Match all collected text content (from speech-to-text transcription and OCR extraction) against the pre-built violation warning keyword database.

Finally, a Multi-Criteria Decision Analysis (MCDA) model is applied to analyze and calculate the violation statements made by the streamers during their live broadcasts, to derive the final Violation Warning Score and determine whether to trigger an early warning.

The Violation Warning Score is calculated using the following formula:
Violation Warning Score = (Number of hits for single 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 live stream room in the past month * 0.1)

The early warning judgment rules are as follows: No early warning prompt will be triggered when the calculated Violation Warning Score is ≤ 1, while an early warning prompt will be triggered when the score exceeds 1.

搜集汇总

数据集介绍

特点

该数据集用于监测快手平台上酒类直播带货中的违法行为，通过分析直播内容中的关键词和组合关键词，计算违法预警值，并根据预设规则触发预警。数据集包含1319条记录，每季度更新一次，主要应用于规范企业快手直播行为。