The era of big data has brought rapid growth and widespread application of data, but the imperfections in the existing data integration system have become obstacles to its high-quality development. The conflict between data security and shared utilization is significant, with traditional data integration methods risking data leakage and privacy breaches. The proposed Privacy-Preserving Record Linkage (PPRL) technology, has effectively resolved this contradiction, enabling efficient and secure data sharing. Currently, many solutions have been developed for PPRL issues, but existing assessments of PPRL methods mainly focus on single indicators. There is a scarcity of comprehensive evaluation and comparison frameworks that consider multiple indicators of PPRL(such as linkage quality, computational efficiency, and security), making it challenging to achieve a comprehensive and objective assessment. Therefore, it has become an urgent issue for us to conduct a multi-indicator comprehensive evaluation of different PPRL methods to explore the optimal approach. This article proposes the use of an modified CRITIC method to comprehensively evaluate PPRL methods, aiming to select the optimal PPRL method in terms of linkage quality, computational efficiency, and security. The research results indicate that the improved CRITIC method based on mathematical statistics can achieve weight allocation more objectively and quantify the allocation process effectively. This approach exhibits exceptional objectivity and broad applicability in assessing various PPRL methods, thereby providing robust scientific support for the optimization of PPRL techniques.