宠物食品类快手直播带货违法监测预警数据
收藏资源简介：
对采集的快手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 ≤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 conducts transcription and analysis on livestream video content of Kuaishou platform influencers who live stream pet food products, and processes and analyzes the oral speech content of these influencers. It counts the frequency and number of violations of pre-defined sensitive violation keywords (including but not limited to "absolutely safe", "prevent", "diarrhea", "skin disease", "parasite", "enhance immunity", "no additives", "all natural", "treat", "money-back guarantee", "no side effects", "definitely effective") during livestreams, and issues warnings or takes corresponding actions based on trigger rules. This work provides data support for the Market Supervision Administration in its jurisdiction to standardize the live streaming behaviors of enterprises on Kuaishou. The collected livestream videos are preprocessed via the following five steps: 1. Segment the original video files into clips with a maximum duration of 10 minutes per segment. 2. Convert the segmented video content to audio (i.e., extract audio from the clips). 3. Perform speech-to-text (STT) conversion on the extracted audio data. 4. Extract text from images captured from the original videos using Optical Character Recognition (OCR) technology. 5. Match the collected text content against the pre-established illegal warning keyword database. Finally, a Multi-Criteria Decision Analysis (MCDA) model is utilized to analyze and calculate the violating statements made by the streamer during the livestream, so as to obtain the illegal warning score and make a warning judgment. Specifically, no warning alert will be triggered when the illegal warning score is ≤ 1, while a warning alert will be triggered when the score is > 1.
