arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2206.12904cs.SI
收藏

我们都在《楚门的世界》里吗?通过自训练识别 Instagram 众包伪造

Are We All in a Truman Show? Spotting Instagram Crowdturfing through Self-Training

Pier Paolo Tricomi, Sousan Tarahomi, Christian Cattai, Francesco Martini, Mauro Conti

AI总结:

本研究提出了首个Instagram众包伪造(CT)互动检测器,通过半监督自训练分类器利用账号特征识别真人参与的虚假互动,在1293个CT样本上达95% F1分数,并发现20位超级网红超20%互动为人为制造。

摘要:

网红营销在2022年创造了160亿美元的收入。通常,网红越受欢迎,其合作报酬就越高。因此,出现了许多通过机器人或虚假账号来提升个人资料受欢迎程度指标的服务。然而,最近真人开始使用其真实账号参与此类提升活动以获取金钱奖励,生成了极难检测的非真实内容。迄今为止,尚无研究尝试在Instagram上检测这种被称为众包伪造(CT)的新现象。在本研究中,我们提出了首个Instagram CT互动检测器。我们的算法通过半监督学习利用个人资料特征来识别参与CT活动的账号。与迄今为止用于识别虚假账号的监督方法相比,半监督模型可以利用大量未标记数据来提高性能。我们从11家提供商处购买并研究了1293个CT个人资料,以构建我们的自训练分类器,该分类器达到了95%的F1分数。我们在真实环境中测试了我们的模型,通过检测和分析20位超级网红(即拥有超过一百万粉丝的网红)的CT互动,发现超过20%的互动是人为制造的。我们分析了CT个人资料和评论,表明仅根据其生成的内容很难检测这些活动。

英文摘要:

Influencer Marketing generated $16 billion in 2022. Usually, the more popular influencers are paid more for their collaborations. Thus, many services were created to boost profiles' popularity metrics through bots or fake accounts. However, real people recently started participating in such boosting activities using their real accounts for monetary rewards, generating ungenuine content that is extremely difficult to detect. To date, no works have attempted to detect this new phenomenon, known as crowdturfing (CT), on Instagram. In this work, we propose the first Instagram CT engagement detector. Our algorithm leverages profiles' characteristics through semi-supervised learning to spot accounts involved in CT activities. Compared to the supervised approaches used so far to identify fake accounts, semi-supervised models can exploit huge quantities of unlabeled data to increase performance. We purchased and studied 1293 CT profiles from 11 providers to build our self-training classifier, which reached 95\% F1-score. We tested our model in the wild by detecting and analyzing CT engagement from 20 mega-influencers (i.e., with more than one million followers), and discovered that more than 20% was artificial. We analyzed the CT profiles and comments, showing that it is difficult to detect these activities based solely on their generated content.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑