Using probabilistic generative models for ranking risks of Android apps

By Hao Peng, Christopher S. Gates, Bhaskar Pratim Sarma, Ninghui Li, Yuan (Alan) Qi, Rahul Potharaju, Cristina Nita-Rotaru, Ian M. Molloy. 2012

421 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Information Systems, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Information Systems

In Signal Processing

Citations in the corpus

Yes, Machine Learning Can Be More Secure! A Case Study on Android Malware Detection (2017, background): “DURING the last decade, machine learning has been increasingly applied in security-related tasks, in response to the increasing variability and sophistication of modern attacks [1], [3], [6], [27], [33].”
DREBIN: Effective and Explainable Detection of Android Malware in Your Pocket (2014, methodology): “Experiments with 123,453 applications from different markets and 5,560 recent malware samples demonstrate the efficacy of our method: DREBIN outperforms related approaches [13, 26, 33] as well as 9 out of 10 popular antivirus scanners.”
Apposcopy: semantics-based detection of Android malware through static analysis (2014, background): “Many recent malware detectors, such as [8, 22, 10, 33, 6], use machine learning to detect zero-day malware.”
CopperDroid: Automatic Reconstruction of Android Malware Behaviors (2015, background): “Such an understanding is an interesting starting point for designing techniques to enforce security policies [42] and avoid the installation of apps requesting a dangerous combination [16] or an overprivileged set of permissions [18], [33].”
DroidAPIMiner: Mining API-Level Features for Robust Malware Detection in Android (2013, methodology): “[17] employ probabilistic generative models to compute a real risk score of Android apps based on the permissions that they request.”
Vetting undesirable behaviors in android apps with permission use analysis (2013, methodology): “Probabilistic models of permission request patterns [32] or permission request sets [47] were also used to indicate the risk of new applications.”
A Machine-learning Approach for Classifying and Categorizing Android Sources and Sinks (2014, methodology): “[46] successfully used various machine-learning approaches to detect malicious Android applications.”
DroidChameleon: evaluating Android anti-malware against transformation attacks (2013, background): “Many works have been done towards discovery and characterization of smartphone malware [7, 15, 20, 23, 24, 29, 30].”
WHYPER: Towards Automating Risk Assessment of Mobile Applications (2013, background): “Previous work has looked at permissions [2–5], code [6–10], and run-time behavior [11–13].”
Structural detection of android malware using embedded call graphs (2013, methodology): “al in [31], where we implement the latter using”
FeatureSmith: Automatically Engineering Features for Malware Detection by Mining the Security Literature (2016, background): “Peng et al. applied a Naive Bayes model to assess how risky apps are given the permissions they request [35].”
Semantics-Aware Android Malware Classification Using Weighted Contextual API Dependency Graphs (2014, neutral): “D.2.4 [Software Engineering]: Software/Program Verification— Validation; D.4.6 [Operating Systems]: Security and Protection— Invasive software”
A Probabilistic Discriminative Model for Android Malware Detection with Decompiled Source Code (2015, methodology): “AUC of ROC is a popular choice used by many researchers [3], [8].”
A Survey of App Store Analysis for Software Engineering (2017, background): “[181] proposed an app risk rating system trained on metadata from name, category and set of permissions.”
Following Devil's Footprints: Cross-Platform Analysis of Potentially Harmful Libraries on Android and iOS (2016, background): “…PHAs, most of which aims at detecting the code [37], [38], [39], [40], [41], [42], [43], [44], [45] for performing potentially harmful behaviors (e.g., stealing the user’s sensitive information), and protecting the Android system from various attacks [46], [47], [48], [49], [50], [51], [52], [53].”
MARVIN: Efficient and Comprehensive Mobile App Classification through Static and Dynamic Analysis (2015, background): “[25] were the first to propose risk scoring for Android apps to improve the communication of the risk that comes with installing an app to users.”