Deceptive Patterns (also known as Dark Patterns) are manipulative interface elements that can cause users to experience financial, temporal, and privacy-related losses. While Deceptive Patterns have been extensively studied in Android applications, their prevalence in iOS apps remains largely unexplored, despite significant ecosystem differences and iOS's growing popularity among younger users. Notably, Apple's tight control over its ecosystem has fostered public perception of iOS being the safer platform and as a byproduct, iOS users may be less vigilant towards app-related risks. To investigate how the prevalence of Deceptive Patterns on iOS compares to Android, we conducted a review of the same 143 mobile apps across both platforms. Our analysis reveals statistically significant differences between Deceptive Patterns on iOS and Android, with iOS apps exhibiting more instances overall (1477 vs. 1398). The findings suggest that iOS users may be more vulnerable to the risks posed by Deceptive Patterns. Furthermore, our analysis identified four specific types of Deceptive Patterns with notable discrepancies between the mobile platforms, indicating potential influences by app store guidelines and developer tools, and the rise of A/B testing Deceptive Patterns. These findings highlight the need to further explore different digital platforms and user protections on mobile devices.

Recommendation systems play a vital role in many online platforms, with their primary objective being to satisfy and retain users. As directly optimizing user retention is challenging, multiple evaluation metrics are often employed. Current methods often use multi-task learning to optimize these measures. However, they usually miss that users have personal preferences for different tasks, which can change over time. Identifying and tracking the evolution of user preferences can lead to better user retention. To address this issue, we introduce the concept of “user lifecycle,” consisting of multiple stages characterized by users’ varying preferences for different tasks. We propose a novel Stage-Adaptive Network (STAN) framework for modeling user lifecycle stages. STAN first identifies latent user lifecycle stages based on learned user preferences and then employs the stage representation to enhance multi-task learning performance. Our experimental results using both public and industrial datasets demonstrate that the proposed model significantly improves multi-task prediction performance compared to state-of-the-art methods, highlighting the importance of considering user lifecycle stages in recommendation systems. Online A/B testing reveals that our model outperforms the existing model, achieving a significant improvement of 3.05% in staytime per user and 0.88% in CVR. We have deployed STAN on all Shopee live-streaming recommendation services.

Outstanding users (OUs) denote the influential, "core" or "bridge" users in the online community. How to accurately detect and rank them is an important problem for third-party online service providers and researchers. Conventional efforts, ranging from early graph-based algorithms to recent machine learning-based approaches, typically rely on an entire network's information or at least ego networks. However, for privacy-conscious users or newly-registered users, such information is not easily accessible. To address this issue, we present DeepPick, a novel framework that considers both the generalization and specialization in the detection task of OUs. For generalization, we introduce deep neural networks to capture nonlinear features. For specialization, we leverage the traditional well-defined metrics to preserve common features. Extensive experiments based on real-world datasets demonstrate that our approach achieves a high efficacy in terms of detection performance against the state-of-the-art.

Text-based passwords have long acted as the dominating authentication method. Leet, as one of the significant components in password, has not been paid enough attention yet. In this paper, we systematically study the presence of Leet in passwords. We define single and pattern forms of Leet and propose a matching approach to check whether a user password contains Leet. We extract the most prevalent counterpart pairs of Leet manifestations. Afterward, we examine the effect of Leet in passwords by incorporating Leet transformation into the probabilistic context-free grammar(PCFG) method to crack passwords. We construct the first comprehensively analyzed dictionary of Leets for passwords, which is confirmed suitable for most datasets by user survey. Experiments on four leaked password sets demonstrate that distinguished Leet usage accumulates to account for around 1% of the total dataset. Only 5% of high-frequency Leets replacement could increase the cracking rate by 0.55%. For crackers, incorporating popular Leets aids to improve password cracking performance. For users, adopting low-frequency Leets could strengthen their passwords. This research provides a new perspective to investigate Leet transformations in passwords.

It is arguable that academics working in accessibility want to improve lives. Last year marked 30 years since the first accessible computing publication [46] on supporting people with Color Vision Deficiency (CVD - commonly called colorblindness), making it a great time to review the CVD academic research that has been completed since then and assess its broad availability to the general population. In this paper, we identify seven central themes in the CVD academic literature, and employ a novel technique (PDI - Persona Driven Inquiry) to explore if and how that literature is available to the general public. Subject to some limitations, we found that this research is indeed available.