化妆品类微信直播带货违法监测预警数据
收藏资源简介：
对采集的微信视频号平台带货品类为化妆品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：祛斑美白、防晒、防脱发、祛痘、滋养、修护、抗皱、紧致、舒缓、控油、去角质等）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为微信视频号直播行为，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 livestream video content from cosmetics category influencers on the WeChat Video Channels platform, with processing and analysis conducted on the oral speech content of the influencers. Based on the frequency and count of violations of pre-set sensitive violation keywords (including spot lightening and whitening, sunscreen, anti-hair loss, anti-acne, nourishing, repairing, anti-wrinkle, firming, soothing, oil-control, exfoliating, etc.) by the influencers during the livestream, warnings or penalties will be issued in accordance with the trigger condition rules. This work provides data support for regulating the livestreaming behaviors of enterprises within the jurisdiction of the Zhuji City Market Supervision Administration on the WeChat Video Channels platform. The preprocessing of the collected livestream videos is carried out in five steps: 1. Segment the original video files into clips with a maximum duration of 10 minutes each; 2. Convert the segmented video content into speech audio; 3. Perform speech-to-text conversion on the segmented clips that have completed video-to-speech conversion; 4. Extract text from images captured from the original videos using Optical Character Recognition (OCR) technology; 5. Match the obtained text content against the violation warning keyword database. Finally, a multi-criteria decision analysis (MCDA) model is applied to analyze and calculate the violating statements made by the streamer during the livestream, to derive the violation warning score and the warning determination result. The violation warning score is calculated as follows: Violation Warning Score = (Number of single violation keyword hits * 0.25) + (Number of combined violation keyword hits * 0.3) + (Number of combined warning keyword hits identified via image recognition * 0.35) + (Number of historical violation records of the livestream room in the past month * 0.1) The final violation warning score is calculated using the above formula. No warning prompt is triggered when the score ≤ 1, while a violation warning prompt is triggered when the score > 1.
