Abstract: For the past years, governments have paid more and more attention to major engineering projects, but major engineering projects have the characteristics of long cycle and much difficulty, and there is a great degree of uncertainty in the construction process, which makes the identification and analysis of risk factors more complex. It needs to be noted that the resources of the project construction entity are limited. Distinct control measures should be implemented for risk factors with varying important degrees. Under such circumstances, to allocate the enterprise resources reasonably, the project team must choose a scientifically and rationally sound decision-making method to identify the risk factors with higher degrees. Therefore, it is of great practical significance to strengthen the identification and analysis of risk factors in major engineering projects.
To effectively improve the level of risk management of major engineering projects, this study aims to analyze the identification and evaluation of risk factors from the perspective of public concern by combining an online-comment analysis and a large-scale group decision making method. Firstly, considering the occurrence of safety incidents of major engineering projects will have a very large negative impact on social stability, the public has a high level of concern about the incidents and risks, the traditional methods have great limitations in dealing with the hot spots of public concern, and therefore, online comments related to major engineering risk factors from micro blogs are extracted and analyzed by using web crawler and content mining technology. Based on this, five first-level risk factors reflecting hot spots are determined. Secondly, with the rapid development of electronic technology, more decision-makers can easily participate in the assessment process of risk factors, and the analytical conclusions obtained are more in line with the real situation. The project team selects 100 decision makers to participate in the analysis process of risk factors, and the preference information is converted into interval 2-tuple linguistic phrases. Thirdly, the consensus building process based on K-means clustering method is used to obtain the preference information of subgroups. Finally, the influence degree of each risk factor is determined based on IVTWA operator. The effectiveness of the proposed method is verified by the risk factor analysis of a hydropower project.
This paper focuses on the identification and analysis of risk factors from the public perspective. In the future research, it is necessary to further explore the risk factor analysis methods from multiple perspectives, and then more reasonable analysis conclusions can be obtained.

Key words: major engineering project, risk factor, content mining, group decision-making approach

摘要： 为了有效提升重大工程项目风险管理的水平,本研究旨在基于内容挖掘技术和群决策理论通过在线评论和群体参与来研究风险因素的识别和分析问题。首先,利用网络爬虫和内容挖掘技术,从微博平台上提取与重大工程风险因素相关的在线评论并进行分析,基于此确定了体现关注热点的五个一级风险因素。其次,项目组选取了100名决策人员参与风险因素的分析问题,并将这些人员的偏好信息转换为二元语言短语。再次,采用基于K-均值聚类的共识达成方法获取子群体的偏好信息;最后,基于IVTWA算子确定各项风险因素的影响程度。通过某水电工程的风险因素分析验证了所提方面的有效性。

关键词: 重大工程项目, 风险因素, 内容挖掘, 群决策