Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2410.00277

Towards Precise Detection of Personal Information Leaks in Mobile Health Apps

Alireza Ardalani·Joseph Antonucci·Iulian Neamtiu
View original source ↗Read source document ↗Export metadata ↓

Abstract

Mobile apps are used in a variety of health settings, from apps that help providers, to apps designed for patients, to health and fitness apps designed for the general public. These apps ask the user for, and then collect and leak a wealth of Personal Information (PI). We analyze the PI that apps collect via their user interface, whether the app or third-party code is processing this information, and finally where the data is sent or stored. Prior work on leak detection in Android has focused on detecting leaks of (hardware) device-identifying information, or policy violations; however no work has looked at processing and leaking of PI in the context of health apps. The first challenge we tackle is extracting the semantic information contained in app UIs to discern the extent, and nature, of personal information. The second challenge we tackle is disambiguating between first-party, legitimate leaks (e.g,. the app storing data in its database) and third-party, problematic leaks, e.g., processing this information by, or sending it to, advertisers and analytics. We conducted a study on 1,243 Android apps: 623 medical apps and 621 health&fitness apps. We categorize PI into 16 types, grouped in 3 main categories: identity, medical, anthropometric. We found that the typical app has one first-party leak and five third-party leaks, though 221 apps had 20 or more leaks. Next, we show that third-party leaks (e.g., advertisers, analytics) are 5x more frequent than first-party leaks. Then, we show that 71% of leaks are to local storage (i.e., the phone, where data could be accessed by unauthorized apps) whereas 29% of leaks are to the network (e.g., Cloud). Finally, medical apps have 20% more PI leaks than health&fitness apps, due to collecting additional medical PI.

Explore related subjects

cs.CRcs.SE

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Alireza Ardalani, Joseph Antonucci, Iulian Neamtiu. 2024-09-30. Towards Precise Detection of Personal Information Leaks in Mobile Health Apps. https://arxiv.org/abs/2410.00277

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Empirical Evaluation of Membership Inference Attacks on NLP Text Classifiers: A Baseline Study on SST-2

Membership inference attacks (MIAs) try to determine whether a specific record was used to train a model, a privacy risk that matters in natural language processing (NLP), where training data can contain sensitive user text. This paper presents a controlled benchmark of membership inference vulnerability for text classification on the GLUE SST-2 sentiment dataset. A TF-IDF + Logistic Regression pipeline and a fine-tuned DistilBERT classifier are compared under a loss-threshold MIA, with utility measured by development accuracy and macro F1. DistilBERT reached 0.9466 accuracy and 0.9460 macro F1 against 0.8756 and 0.8727 for Logistic Regression, yet both models leaked membership signal (Attack AUC 0.5615 and 0.5800, respectively). Two mitigations were tested. Stronger regularization reduced leakage for Logistic Regression at a visible utility cost, whereas fine-tuning DistilBERT for 2 epochs instead of 3 reduced leakage with negligible accuracy loss. Lightweight training adjustments can improve the privacy-utility trade-off without complex defenses.

cs.CR↗
arXiv2026

Empirical Evaluation of Data Poisoning Attacks in Supervised Learning

Data poisoning corrupts training data to degrade a model or to plant attacker-controlled behavior. This study evaluates two representative training-time attacks, label flipping and backdoor poisoning, on MNIST and Fashion-MNIST with three baseline classifiers: Logistic Regression, Linear SVM, and Random Forest. Clean training is compared with poisoning rates of 5%, 10%, and 20% using clean-test accuracy, macro-precision, macro-recall, macro-F1, and, for backdoors, attack success rate. Label flipping caused clear degradation, largest for Logistic Regression and Linear SVM, while Random Forest stayed comparatively stable. Backdoor poisoning reached attack success rates from 0.9667 to 1.0000 on both datasets and all three models while often keeping clean-test performance near baseline. The results separate indiscriminate poisoning, which shows up in standard metrics, from targeted backdoor poisoning, which stays comparatively stealthy while embedding highly effective malicious behavior, and they support security-oriented evaluation beyond conventional clean-test metrics.

cs.CR↗
arXiv2026

DeFiFusion: Combining Transaction Events with Smart Contracts to Detect Price Manipulation Attacks

Decentralized Finance (DeFi) has emerged as a rapidly growing blockchain-based financial service, where market transaction dynamics and underlying smart contract logic are intricately intertwined. This autonomous interplay, while eliminating centralized intermediaries, significantly expands the vulnerability surface of DeFi protocols to Price Manipulation Attacks (PMAs), which have already inflicted catastrophic financial losses. Despite their gravity, existing detection paradigms suffer from fundamental limitations. Transaction-centric methods lack awareness of contract execution semantics, making them prone to false positives under legitimate market volatility, while static contract analyses ignore real transaction behaviors and frequently report vulnerabilities that are infeasible to exploit in practice. We present DeFiFusion, a dual-modal PMA detection framework that closes this gap by jointly modeling transaction events and smart contract semantics within a unified pipeline. Our core insight is that PMA maliciousness emerges only from the interaction between transaction behaviors and the contract logic they exploit; neither signal suffices in isolation. Accordingly, we derive price-manipulation-aware event encoding for extracting fine-grained temporal and economic features tailored to manipulation patterns. We further introduce LLM-based contract semantic extraction to supply the execution-logic context that prior behavioral methods lack. To fuse these modalities, we propose a Dual-Modal Projection-Fusion Transformer with T5-style relative positional encoding, capturing the cyclic multi-stage execution structures that distinguish PMAs from benign market activity. Extensive experiments demonstrate that DeFiFusion consistently achieves state-of-the-art detection performance, effectively recalling 222 of the 225 PMA cases while maintaining a precision of 96.10%.

cs.CR↗