宠物食品类抖音直播带货违法监测预警数据

对采集的抖音平台带货品类为宠物食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：绝对安全、预防、腹泻、皮肤病、寄生虫、增强免疫力、无添加、纯天然、治疗、无效退款、没有副作用、肯定有用等）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抖音直播行为，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 is developed based on the processing and analysis of influencer live stream videos in the pet food category collected from the Douyin platform. The spoken language content of the influencers during live broadcasts is first processed and analyzed. Warnings or corresponding treatments will be issued based on the triggering condition rules, according to the frequency and number of violations of pre-set sensitive prohibited keywords by the influencers in live streams. The pre-set sensitive prohibited keywords include but are not limited to: absolute safety, prevention, diarrhea, skin diseases, parasites, enhance immunity, additive-free, 100% natural, treatment, money-back guarantee, no side effects, definitely effective, etc. This work provides data support for market supervision bureaus in their jurisdiction to standardize the live streaming behaviors of enterprises on the Douyin platform. The collected live stream videos undergo the following preprocessing steps: Step 1: Slice the original video files into chunks with a maximum duration of 10 minutes each. Step 2: Convert the sliced videos into audio files. Step 3: Perform speech-to-text conversion on the sliced audio files obtained from the previous step. Step 4: Extract text from images captured in the original videos using OCR technology. Step 5: Match all the extracted text content against the violation warning keyword database. Finally, a multi-criteria decision analysis model is used to analyze and calculate the violation statements made by the streamer during the live broadcast, so as to obtain the violation warning score and the judgment of whether to trigger a warning. The formula for calculating the final violation warning score is as follows: Violation Warning Score = (Number of hits for single prohibited keywords * 0.25) + (Number of hits for combined prohibited keywords * 0.3) + (Number of combined warning keywords identified via image recognition * 0.35) + (Number of historical violation records of the streamer in the past month * 0.1) The warning triggering rules are: If the final Violation Warning Score is ≤ 1, no warning alert will be triggered; if the Score is > 1, a violation warning alert will be triggered.