基于Apriori的装配式建筑质量影响因素分析

Impact Factors Analysis on Prefabricated Building Quality Based on Apriori Algorithm

摘要:

采用数据挖掘的方法，在装配式建筑质量的文献中提取影响因素，利用Apriori算法得到筛选后的关键指标并分析。研究发现，基于大数据能够减少人为因素对指标选取的干扰，并通过寻找影响因素和装配式建筑质量的相关关系，可得到更具普遍性的规律。在5M1E分析法的影响因素中，“法”的影响因素支持度较大，表明该板块包含的影响因素对装配式建筑的质量有更显著的影响。通过置信度分析，提出使用强关联规则分析，并发现“构件生产企业的标准化程度”不仅具有较高的支持度，在关联因素中出现的频率也最高，从而提出建议并为合理评价质量影响因素及制定管理措施提供有效参考。

Abstract:

Through data mining， quality impact factors were extracted from prefabricated building articles， then filtered into key factors and analyzed by the Apriori algorithm. Results indicate that big data help to reduce the interference on index selection caused by human factors， while a more universal law could be obtained by observing the correlation between impact factors and the quality of prefabricated buildings. Among all 5M1E factors， the “Method” branch gains greater support， indicating a significant impact on prefabricated buildings’ quality. Based on confidence analysis and the proposed strong association rules， it is found that “standardization of component manufacturers” receives not only a higher support degree but also the greatest frequency among the associated factors. Suggestions are put forward and shall be an effective reference for quality impact factors’ evaluation and management formulation.