基于聚类分析的人工智能产品发展风险评估方法  

The risk assessment method of artificial intelligence product development based on cluster analysis

作　　者：袁炳夏[1] YUAN Bingxia(Huizhou University,Huizhou Guangdong 516007,China)

机构地区：[1]惠州学院,广东惠州516007

出　　处：《智能计算机与应用》2021年第7期177-179,共3页Intelligent Computer and Applications

摘　　要：人工智能产品已经多元化地应用于社会生活中,推动技术发展进步的同时也带来了一定的安全风险,准确地评估人工智能产品发展风险可以为其可持续发展提供良好的参考与借鉴。为此,提出基于聚类分析的人工智能产品发展风险评估方法。首先,以主题词的形式设计人工智能产品目前应用范围的关联图谱,提取人工智能产品发展的关键特征;其次,依据产品发展特征深度提取产品发展中应用的高频主题词,结合发展过程中可能存在的风险因素,对其进行聚类处理,实现人工智能产品发展的风险评估。实验发现,人工智能产品发展确实存在一定的风险,所设计的聚类分析方法可以准确评估其发展风险。Artificial intelligence products have been widely used in social life,promoting the development and progress of technology,but also brings a certain security risk.Accurate assessment of the development risk of artificial intelligence products can provide a good reference for its sustainable development.Therefore,a risk assessment method of AI product development based on cluster analysis is proposed.Firstly,the association map of the current application range of artificial intelligence products is designed in the form of subject words,and the key features of the development of artificial intelligence products are extracted.Secondly,the high-frequency subject words applied in the development of products are deeply extracted according to the characteristics of product development,and combined with the possible risk factors in the development process,they are clustered to realize the risk assessment of the development of artificial intelligence products.Experiments show that there are some risks in the development of AI products,and the cluster analysis method can accurately evaluate the development risk.

关 键 词：聚类分析 人工智能 聚类结果 风险评估 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]