I-Corps: Market Impact Identification of Dyadic Attribution Model for Disposition Assessment Using Online Games

I-Corps：使用在线游戏进行处置评估的二元归因模型的市场影响识别

• 批准号: 1505195
• 负责人: Shuyuan Metcalfe
• 金额: $ 5万
• 依托单位: Florida State University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-12-15 至 2016-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1505195&HistoricalAwards=false
• 关键词: Corps; Market; Impact; Identification; Dyadic

Both public and private agencies have a need to identify psychological attributes that can lead to potentially undesired behavior or best-fit in the organization. Communities across the US and throughout the world are looking for new approaches to protect againstpotentially dangerous behaviors. Cyber threats specific to our military, privacy, personal safety, financial security, and national security are initiated by human factors. Language used reveals the tendency of a user's behavior, which provides a window into the potential stability, dependability, and trustworthiness of an individual. The ability to identify potentially harmful human behavior before it is acted out becomes increasingly viable. The need for increased intervention measures by appropriate authorities before undesired behavior surfaces must be addressed. By measuring trustworthiness, organizational insider threats can be mitigated. By screening military personnel, stress-related disorders can be identified early on. By understanding students' disposition, cyber-bullying behavior can be recognized in order to safeguard students. By paying attention to the disposition of law enforcement personnel, community interactions can be less contentious and more effective. This I-Corps team has developed a methodology to identify threatening human disposition and deceptive motivation through analysis of online communication behaviorThis technology has the ability to extract actors' language-action features based on their online communication to establish dyadic attribution models of trustworthiness. During interactive game playing, the technology analyzes online actors' choices based on their online conversations, to discern their communicative intent. The language-action features will feed a causality reasoning system wherethe trustworthiness of the social actor is analyzed and updated in a recursive Bayesian inference framework. The analysis from this dyadic attribution model has the potential for assessing threatening human disposition and intent (e.g., disgruntled) based on information behavior. The further utilization of this model is to apply in similar instances such as identifying actors with tendencies for suicide, murder,rage, and violence based on their online behavior. With this product, the authorities can analyze psychological stability and respond when necessary to provide counseling so that tragic consequences can

公共和私人机构都需要确定可能导致潜在不良行为或最适合组织的心理属性。美国和世界各地的社区都在寻找新的方法来防止潜在的危险行为。针对我们的军事、隐私、个人安全、金融安全和国家安全的网络威胁是由人为因素引发的。使用的语言揭示了用户的行为倾向，这为了解一个人的潜在稳定性、可靠性和可信度提供了一个窗口。在行为发生之前识别潜在有害的人类行为的能力变得越来越可行。必须解决有关当局在不良行为出现之前增加干预措施的必要性。通过衡量可信度，可以减轻组织内部的威胁。通过对军人的筛查，可以早期发现与压力相关的障碍;通过了解学生的性格，可以识别网络欺凌行为，以保护学生。通过关注执法人员的部署，社区互动可以减少争议，提高效率。本研究团队开发了一种通过分析在线交流行为来识别威胁性人格倾向和欺骗性动机的方法，该技术能够提取行为者基于在线交流的语言-行为特征，建立可信度的二元归因模型。在互动游戏中，该技术根据在线玩家的在线对话分析他们的选择，以辨别他们的交流意图。语言动作特征将为因果推理系统提供信息，其中社会行为者的可信度在递归贝叶斯推理框架中进行分析和更新。来自这种二元归因模型的分析具有评估威胁性人类倾向和意图的潜力（例如，不满意的）基于信息行为。该模型的进一步利用是适用于类似的情况下，如识别行为人的自杀倾向，谋杀，愤怒，暴力基于他们的在线行为。有了这个产品，当局可以分析心理稳定性，并在必要时作出反应，提供咨询，以便悲剧性的后果可以