Toma Anca Mirela, Cerchiello Paola
Department of Economics and Management Science, University of Pavia, Pavia, Italy.
Front Artif Intell. 2020 Apr 16;3:18. doi: 10.3389/frai.2020.00018. eCollection 2020.
Initial coin offerings (ICOs) are one of the several by-products in the world of the cryptocurrencies. Start-ups and existing businesses are turning to alternative sources of capital as opposed to classical channels like banks or venture capitalists. They can offer the inner value of their business by selling "tokens," i.e., units of the chosen cryptocurrency, like a regular firm would do by means of an IPO. The investors, of course, hope for an increase in the value of the token in the short term, provided a solid and valid business idea typically described by the ICO issuers in a white paper. However, fraudulent activities perpetrated by unscrupulous actors are frequent and it would be crucial to highlight in advance clear signs of illegal money raising. In this paper, we employ statistical approaches to detect what characteristics of ICOs are significantly related to fraudulent behavior. We leverage a number of different variables like: entrepreneurial skills, Telegram chats, and relative sentiment for each ICO, type of business, issuing country, team characteristics. Through logistic regression, multinomial logistic regression, and text analysis, we are able to shed light on the riskiest ICOs.
首次代币发行(ICO)是加密货币领域的几种副产品之一。初创企业和现有企业正在转向传统渠道(如银行或风险投资家)以外的其他资本来源。它们可以通过出售“代币”(即所选加密货币的单位)来展示其业务的内在价值,就像普通公司通过首次公开募股(IPO)那样。当然,投资者希望代币价值在短期内上涨,前提是ICO发行者在白皮书中通常会描述一个可靠且有效的商业理念。然而,不法行为者的欺诈活动屡见不鲜,提前突出非法募资的明显迹象至关重要。在本文中,我们采用统计方法来检测ICO的哪些特征与欺诈行为显著相关。我们利用了许多不同的变量,如:创业技能、Telegram聊天记录、每个ICO的相对情绪、业务类型、发行国家、团队特征。通过逻辑回归、多项逻辑回归和文本分析,我们能够揭示风险最高的ICO。