The fast growth of Social Networks (SNs), amplified by the ever-increasing use of smartphones, has intensified online cybercrimes. This trend has accelerated digital investigations through SNs. In particular, camera Sensor Pattern Noise (SPN) uniquely characterizing each smartphone has attracted a lot of attention. In this paper, we propose a clustering and classification approach to achieve Smartphone Identification (SI) and User Profiles Linking (UPL) across SNs to provide investigators with significant findings in SN forensics. We test the proposed methods on a dataset of 2,000 images shared on Google+, Facebook, WhatsApp, and Telegram taken by 10 smartphones. The results show the effectiveness of our approach in distinguishing between the same models of the same smartphone brands despite the loss of image detail through the compression process on SNs. The average of sensitivity and specificity values are, respectively, 98.5% and 99.5% for SI and UPL across the SNs.
Social network forensics through smartphones and shared images / Rouhi, R.; Bertini, F.; Montesi, D.; Li, C. -T.. - ELETTRONICO. - (2019), pp. 8739237.1-8739237.6. (Intervento presentato al convegno 7th International Workshop on Biometrics and Forensics, IWBF 2019 tenutosi a Messico nel 2019) [10.1109/IWBF.2019.8739237].
Social network forensics through smartphones and shared images
Bertini F.;
2019-01-01
Abstract
The fast growth of Social Networks (SNs), amplified by the ever-increasing use of smartphones, has intensified online cybercrimes. This trend has accelerated digital investigations through SNs. In particular, camera Sensor Pattern Noise (SPN) uniquely characterizing each smartphone has attracted a lot of attention. In this paper, we propose a clustering and classification approach to achieve Smartphone Identification (SI) and User Profiles Linking (UPL) across SNs to provide investigators with significant findings in SN forensics. We test the proposed methods on a dataset of 2,000 images shared on Google+, Facebook, WhatsApp, and Telegram taken by 10 smartphones. The results show the effectiveness of our approach in distinguishing between the same models of the same smartphone brands despite the loss of image detail through the compression process on SNs. The average of sensitivity and specificity values are, respectively, 98.5% and 99.5% for SI and UPL across the SNs.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
