The rapid proliferation of websites has posed significant challenges to cyberspace management. To effectively manage websites, we introduce the innovative concept of website association in this paper and propose a website association model, WebAA. Website association seeks to determine whether target websites belong to the same organization based on their features, playing a crucial role in cyberspace management. Unlike website identification, website association enables more accurate organizational alignment through fine-grained analysis of website features. Specifically, malicious website association helps investigators identify the organizations behind them, thereby addressing the root causes of the harm they cause. In simple terms, the proposed website association model in this paper consists of four modules. In the first module, we calculate the similarity score based on external resources by analyzing the dependency relationships of the target websites on these resources. The second module employs the BERT model to analyze the similarity of HTML texts between the target websites and generates a similarity score. The third module focuses on the website domain names, calculating their similarity using Levenshtein distance and obtaining the similarity score. In the fourth module, the final similarity score is obtained by weighting the three similarity scores, which is then used to determine whether the two websites belong to the same organization. Extensive experiments on two real-world datasets demonstrate that our model can efficiently associate thousands of website pairs within milliseconds with an accuracy exceeding 90%. Striving for applicability and replicability, we release ready-to-use raw data from our study.