File sharing anomaly detection in business environments

• Autor(a) principal: Mota, João Manuel Ribeiro
• Data de Publicação: 2018
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/28365
• topic: Network; Anomaly detection; File sharing; Ransomware; Theft; Propagation; Machine learning

File sharing is the activity of making archives (documents, videos, photos) available to other users. Enterprises use file sharing to make archives available to their employees or clients. The availability of these files can be done through an internal network, cloud service (external) or even Peer-to-Peer (P2P). Most of the time, the files within the file sharing service have sensitive information that cannot be disclosed. Equifax data breach attack exploited a zero-day attack that allowed arbitrary code execution, leading to a huge data breach as over 143 million user information was presumed compromised. Ransomware is a type of malware that encrypts computer data (documents, media, ...) making it inaccessible to the user, demanding a ransom for the decryption of the data. This type of malware has been a serious threat to enterprises. WannaCry and NotPetya are some examples of ransomware that had a huge impact on enterprises with big amounts of ransoms, for example WannaCry reached more than 142,361.51$ in ransoms. In this dissertation, we purpose a system that can detect file sharing anomalies like ransomware (WannaCry, NotPetya) and theft (Equifax breach), and also their propagation. The solution consists of network monitoring, the creation of communication profiles for each user/machine, an analysis algorithm using machine learning and a countermeasure mechanism in case an anomaly is detected.