現代生活與網際網路緊密相關，包括網路購物、遊戲、社交互動和資訊檢索等多個方面。這種密不可分的關聯性，使得個人資料在網絡上廣泛流通，更出現了不斷增長的網絡犯罪行為。駭客對於企業和個人的攻擊頻率持續攀升，為普通消費者帶來了嚴重的網路釣魚和詐騙網站的威脅。為了因應此一問題，本研究已成功開發了一個針對繁體中文網頁的安全辨識雛型系統，包括了七個主要模組：使用者介面模組、註冊模組、使用者搜索模組、網路爬蟲模組、專家系統模組、多層感知器模組和使用者回饋模組，同時也設立了三個後臺支援的資料庫。使用者透過搜索欄輸入欲查詢的網址，系統透過詐騙資料庫的網址匹配、專家系統模組和多層感知器模組進行檢測分析，系統最終評估該網址的風險值。期望透過這個雛型系統的建立，能夠提高網民對網絡安全的警覺，並且有效地降低普通網民成為詐騙受害者的風險。

Modern life is closely intertwined with the Internet, encompassing various aspects such as online shopping, gaming, social interaction, and information retrieval. This inherent interconnectivity has led to the widespread circulation of personal data on the web and the proliferation of cybercriminal activities. Hackers continue to escalate their attacks on both enterprises and individuals, posing a grave threat of phishing and fraudulent websites to general consumers. In response to this challenge, our research has successfully developed a prototype system for secure identification of Traditional Chinese web pages. This system comprises seven core modules: the user interface module, registration module, user search module, web crawler module, expert system module, multi-layer perceptron module, and user feedback module. Additionally, three backend support databases have been established. Users can input the web address they wish to investigate into the search bar, and the system leverages the URL matching from the fraud database, along with analysis from the expert system module and multi-layer perceptron module to assess the risk level of the given web address. The aim of establishing this prototype system is to heighten public awareness of online security and effectively reduce the risk of general internet users falling victim to fraud.