A Recommendation System for Privacy in Facebook

[[abstract]]社群網路已成為人們最受歡迎的媒介之一，人們花費許多時間在平台上進行社交互動，與分享近況，大量的個人資料在平台上流通，個人隱私往往在未查覺的情況下被盜用。Facebook為目前最大的社群網站，擁有最多的用戶數，平台上海量的資料也成為網路犯罪者覬覦的目標，網路犯罪者利用僵屍帳號偽裝成一般使用者活躍於社群網路中，藉此竊取個人資料獲取利益。社群網路分享及開放的型態，但也往往忽略了隱私保護的重要性，因此如何讓使用者在享受社群網路所帶來的樂趣之時，同時保護個人隱私是現今值得重視的議題。 本研究針對Facebook的資料可能遭竊取或誤用之行為提出防範與建議機制，利用Facebook API獲取朋友的個人資料，統計各項屬性開放程度，提出相對性隱私機制，做為使用者在隱私設定上的重要參考，並整理使用者的主動發言記錄以及隱私發佈設定，讓使用者能夠檢視過去的發言記錄中是否有不適當的隱私設定，最後利用社群網路的特質提出朋友共同性分析機制，讓使用者能夠重新檢視好友清單中是否為潛在的僵屍帳號。[[abstract]]Social networking has become one of popular communication media, and people take advantage of social networks to share their interest and to interact with acquaintances. A large amount of personal information is available in social networks. This may incur the violation of personal privacy. Currently, Facebook is the most prominent social network and owns the most members. As a result, Facebook is an obvious attacking target for adversaries. The cybercrime using bots that mimic real online social network users and steal personal information to benefit. Sharing and openness are the foundation of an online social network, but the privacy issue is overlooked. How to protect privacy in social networking is an important issue today. In this study, we develop an effective protection and suggestion scheme to help Facebook users to prevent the leak or abuse of their personal information. By the means of Facebook API, we propose a novel relative privacy scheme as an important reference for privacy settings. In addition, our study can assist users to review their detailed privacy settings and to realize information revealed to their friends or the public. Therefore, user are reminded to pay attention to possible privacy threats. Finally, the proposed system can effectively figure out the common features between a user and his/her friends, for example, common hobbies, groups, and mutual friends. This can help users to review their friendships and to possibly find potential socialbots.[[note]]碩