隨著社群網站與共享經濟的興起，Crowdsourcing 資料已在多個領域被廣泛使用。而經由網民提供的喜好資料以可做為產品或服務排序的基礎。但是綜合網民所提供的互相衝突且不齊全的資料獲取正確排序是相當複雜的議題。為了解決這個問題，本研究提出一個新的演算法。本方法修改與增強FCM以達到高可靠性與高使用性。為了驗證可靠性、使用性與排序正確性，本研究並包含一系列的實驗，實驗資料包含真實資料與人造資料。實驗結果顯示本研究較其他方法有較好的可靠性與使用性。而正確性也與最好的Borda Count 不分軒輊.;With the rapid development of social network and online services, crowdsourcing data has been used for many solutions in various fields. The preference sequences obtained through crowdsourcing are valuable resources for ranking. However, the aggregation of incomplete and inconsistent preferences is complicated. To address these challenges, this research proposed a novel method termed robust crowd ranking (RCR) based on a consistent Fuzzy C-means (CFCM) approach to increase the robustness and accessibility of aggregated preference sequences obtained through crowdsourcing. To verify the robustness, accessibility, and accuracy of RCR, comprehensive experiments were conducted using synthetic and real data. The simulation results validated that the RCR outperforms Borda Count, Dodgson, IRV and Tideman methods.
