普通食品类淘宝直播带货违法监测预警数据

对采集的淘宝平台带货品类为普通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：增强免疫力，降血脂，降血糖，抗氧化，改善记忆，缓解视疲劳，促进排铅，清咽，辅助降血压，改善睡眠，促进泌乳，缓解体力疲劳，提高缺氧耐受力，抗辐射，减肥等）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淘宝直播行为，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 performs transcription and analysis on live stream video content of Key Opinion Leaders (KOLs) promoting general food products on the Taobao platform, and processes and analyzes the spoken verbal content of the KOLs during their live broadcasts. Based on the frequency and count of violations of pre-set sensitive prohibited words (such as: enhancing immunity, lowering blood lipid levels, lowering blood glucose levels, anti-oxidation, improving memory, relieving eye fatigue, promoting lead excretion, throat clearing, assisting in lowering blood pressure, improving sleep quality, promoting lactation, relieving physical fatigue, improving hypoxia tolerance, anti-radiation, weight loss, etc.) by the KOLs during live streams, warnings or corresponding treatments are issued in accordance with trigger condition rules. This work provides data support for standardizing the live streaming behaviors of enterprises within the jurisdiction of the Zhuji Municipal Market Supervision Administration.

The collected live stream videos undergo the following preprocessing procedures:
1. Slice the original video files into segments with a maximum duration of 10 minutes each.
2. Convert the video content into audio for the sliced videos.
3. Convert the extracted audio content into text for the processed video slices.
4. Extract text from images captured from the original videos using Optical Character Recognition (OCR) technology.
5. Match the obtained text content against the prohibited early-warning keyword database.

Finally, a multi-criteria decision analysis model is applied to analyze and calculate the prohibited statements uttered by the anchors during live broadcasts, so as to derive the violation early-warning value and make a judgment on whether to trigger an early warning.

The violation early-warning value is calculated using the formula:
Violation Early-warning Value = (Number of hits for single prohibited early-warning keywords * 0.25) + (Number of hits for combined prohibited early-warning keywords * 0.3) + (Number of combined early-warning keywords identified via image recognition * 0.35) + (Number of historical violation records of the live stream room in the past month * 0.1)

According to the calculation result, if the violation early-warning value ≤ 1, no early warning prompt will be triggered; if the value > 1, a violation early warning prompt will be issued.

搜集汇总

数据集介绍

特点

该数据集用于监测淘宝直播中普通食品类带货的违法行为，包含1883条记录，每季度更新一次。通过视频切片、语音转文本、OCR技术等手段提取信息，并与预设的违法关键词库进行匹配，计算出违法预警值，为市场监管部门提供数据支持。