Dynamic collective opinion generation framework for digital transformation barriers analysis in the construction industry

Zhen-Song Chen,Chong-Ze Liang,Ya-Qiang Xu,Witold Pedrycz,Mirosław J. Skibniewski
DOI: https://doi.org/10.1016/j.inffus.2023.102096
IF: 18.6
2023-10-27
Information Fusion
Abstract:The absence of a reliable, dynamic evaluation system has impeded early-stage industrial research progress, particularly in the digital transformation of the construction industry. Moreover, existing studies rarely explore barriers' direct impact on the subject of study and the interplay between barriers. This paper aims to introduce an innovative framework for generating dynamic collective opinions. Dynamic collective opinion generation framework comprises three key components: Collective Opinion Generation (COG), Prediction with Expert Advice (PEA), and Social Network Analysis (SNA). Its goal is to provide dependable decision support when subjective evaluation data from experts is available. Initially, a bi-objective optimization model generates the initial barrier weight vector. The PEA incorporates a loss function to measure the deviation between aggregated PDF and actual observed data, updating the weight vector over time. Next, an influence network covering all barriers is established. Node significance is evaluated through metrics like degree centrality, closeness centrality, and eigenvector centrality. The gravity model based on three metrics is used to determine interrelationships among barriers, resulting in a weight vector capturing these interplays. The two weight vectors are weighted using Nash equilibrium, yielding the ultimate weight vector for barriers. The effectiveness of the proposed dynamic collective opinion generation framework is showcased through a case study on China Construction Third Bureau. Results indicate that talent structure notably influences construction companies' digital transformation. Additionally, market structure and strategic positioning significantly impact digital transformation in this industry.
computer science, artificial intelligence, theory & methods
What problem does this paper attempt to address?