Mobile Commerce - Analysis and Investigation of the Online Safety, Privacy, and Data Forensics of Amazon and Etsy Apps

The COVID19 pandemic has led to the proliferation of the use of online shopping applications among millions of customers worldwide. The enormous potential in technological advancements, particularly mobile technology, has directly impacted mobile commerce, where the shopping process has become so convenient. While the benefits of mobile commerce are multi-fold, the current privacy practices and the extent of user data residue in shopping apps have been less explored. In this paper, we conducted an in-depth, systematic analysis of two of the most popular mobile shopping apps - Amazon and Etsy. Our analysis led to the recovery of user data and shopping activity artifacts from Amazon and Etsy buyer and seller apps on Android/iOS devices. Based on the user data and artifacts found, we have also discussed the implications of default privacy settings, the importance of online safety policies prior to product listings, and implications for research and practice. © 2023 IEEE Computer Society. All rights reserved.

FedTD: Efficiently Share Telemedicine Data with Federated Distillation Learning

With the Internet of Things and medical technology development, patients use wearable telemedicine devices to transmit health data to hospitals. The need for data sharing for public health has become more urgent under the COVID-19 pandemic. Previously, security protection technology was difficult to solve the increasing security risks and challenges of telemedicine. To address the above hindrances, Federated learning (FL) solves the difficulty for companies and institutions to share user data securely. The global server iterative aggregates the model parameters from the local server instead of uploading the user's data directly to the cloud server. We propose a new model of federated distillation learning called FedTD, which allows the different models between local hospital servers and global servers. Unlike traditional federated learning, we combine the knowledge distillation method to solve the non-Independent Identically Distribution (non-IID) problem of patient medical data. It provides a security solution for sharing patients' medical information among hospitals. We tested our approach on the COVID-19 Radiography and COVID-Chestxray datasets to improve the model performance and reduce communication costs. Extensive experiments show that our FedTD significantly outperforms the state-of-the-art. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

The Relationship Between Exercise Habits and Stress Among Individuals With Access to Internet-Connected Home Fitness Equipment: Single-Group Prospective Analysis.

BACKGROUND: Physical activity (PA) confers numerous benefits to health and health care costs, yet most adults are not meeting recommended PA guidelines. Stress may be a factor that influences PA behavior. Research investigating the impact of stress on PA has yielded inconsistent findings. Most studies find that stress negatively impacts PA, but there is some evidence that habitual exercising buffers this association. OBJECTIVE: This study aims to examine the relationship between stress and exercise habits among habitual exercisers with internet-connected home fitness equipment (Peloton Bike) during the COVID-19 lockdown. METHODS: Participants were recruited through Facebook (N=146) and asked to complete an internet-based survey that assessed COVID-19-related stressors, perceived stress associated with those stressors, and general perceived stress. Self-reported exercise was assessed on the survey using the Godin Leisure-time Exercise Questionnaire (GLTEQ). Participants were also asked for consent to access their Peloton usage data through the Peloton platform. From their usage data, the frequency and duration of cycling classes was calculated for 4 weeks prior to and 12 weeks following the survey. Hierarchical regression equations tested the association between stress reported on the survey and subsequent exercise participation. Exercise participation was quantified both as the frequency and duration of Peloton cycling over the 12 weeks following the survey and as self-reported moderate to vigorous activity on a second survey completed by a subset of participants 12 weeks after the initial survey. RESULTS: There were 146 participants in our Peloton analysis sample and 66 in the self-reported exercise analysis. Peloton user data showed that study participants cycled frequently (mean 5.9 times per week) in the month prior to the initial survey, and that presurvey Peloton use was a strong predictor of exercise frequency (R2=0.57; F2,143=95.27; P<.001) and duration (R2=0.58; F2,143=102.58; P<.001) for the 12 subsequent weeks. Self-reported overall exercise likewise showed that this sample was very active, with an average of more than 8 times per week of moderate to vigorous exercise at the initial survey. Self-reported exercise on the initial survey was a strong predictor of self-reported exercise 12 weeks later (R2=0.31; F1,64=29.03; P<.001). Perceived stress did not impact Peloton cycling duration or frequency (P=.81 and .76, respectively) or self-reported exercise (P=.28). CONCLUSIONS: The results suggest that stress did not negatively impact exercise participation among habitually active adults with access to internet-connected home fitness equipment. Habitual exercise may buffer the impact of stress on participation in regular moderate to vigorous activity. Future research should examine the role that the availability of home-based internet-connected exercise equipment may play in this buffering.

CoviFL: Edge-Assisted Federated Learning for Remote COVID-19 Detection in an AIoMT Framework

Detection of COVID-19 has been a global challenge due to the lack of proper resources across all regions. Recently, research has been conducted for non-invasive testing of COVID-19 using an individual's cough audio as input to deep learning models. However, these methods do not pay sufficient attention to resource and infrastructure constraints for real-life practical deployment and the lack of focus on maintaining user data privacy makes these solutions unsuitable for large-scale use. We propose a resource-efficient CoviFL framework using an AIoMT approach for remote COVID-19 detection while maintaining user data privacy. Federated learning has been used to decentralize the CoviFL CNN model training and test the COVID-19 status of users with an accuracy of 93.01 % on portable AIoMT edge devices. Experiments on real-world datasets suggest that the proposed CoviFL solution is promising for large-scale deployment even in resource and infrastructure-constrained environments making it suitable for remote COVID-19 detection. © 2022 IEEE.

Forensic Analysis of Microsoft Teams: Investigating Memory, Disk and Network

Videoconferencing applications have seen a jump in their userbase owing to the COVID-19 pandemic. The security of these applications has certainly been a hot topic since millions of VoIP users’ data is involved. However, research pertaining to VoIP forensics is still limited to Skype and Zoom. This paper presents a detailed forensic analysis of Microsoft Teams, one of the top 3 videoconferencing applications, in the areas of memory, disk-space and network forensics. Extracted artifacts include critical user data, such as emails, user account information, profile photos, exchanged (including deleted) messages, exchanged text/media files, timestamps and Advanced Encryption Standard encryption keys. The encrypted network traffic is investigated to reconstruct client-server connections involved in a Microsoft Teams meeting with IP addresses, timestamps and digital certificates. The conducted analysis demonstrates that, with strong security mechanisms in place, user data can still be extracted from a client’s desktop. The artifacts also serve as digital evidence in the court of Law, in addition to providing forensic analysts a reference for cases involving Microsoft Teams. © 2022, ICST Institute for Computer Sciences, Social Informatics and Telecommunications Engineering.

AWARE: An IoT powered Smart Band with Multi-Tenancy Cardinality

The Pandemic caused due to COVID-19 has overweighed the current healthcare system and made us realize that how unaware we were of our health. Although lots of rational and harsh measured were taken to curb the spread of COVID-19 but still we lost millions of lives. During this pandemic technology played a vital role ranging from the invention of vaccine to remotely monitoring the usage with the help of IoT. Among several emerging technologies wearable smart devices were burgeoning as they are also powered by IoT now. Wearable devices are being used in different scenarios ranging from tracking and monitory infected patients to utilize the data for policy making. Proposed is a framework of an ecosystem "AWARE"which comprises of a smart band with advanced PPG and EEG sensors to detect Heart rate, Heart rate Variability, Respiration rate, SpO2, Step and Sleep data. The sensor data will be transferred to the AWARE application on host mobile through BLE and from mobile the user data it will be transferred to AWARE cloud for pre and post processing using Machine Learning algorithms. AWARE can used for monitoring and detection of health anomalies and diseases such as COVID-19 or chronic lifestyle diseases. AWARE works on a multi-Tenancy cardinality framework were the group (i.e., kids, elders, domestic worker) users can share the cloud storage and receive customized notifications. Also, the group manager (i.e., father, employer) will be notified in case of emergencies. Cloud data can be accessed by the users through dashboards. Government authorities can also access user data through APIs. Wearables are widely accepted these days due to its less intrusiveness. Although some professionals are skeptical of these devices, but the advantages are far beyond the minor pitfalls. In fact, several countries have already implemented wearables devices as the primary medium of COVID-19 detection, monitoring. © 2021 IEEE.

Federal Learning Based COVID-19 Fake News Detection With Deep Self-Attention Network

As social media becomes more and more popular, fake news spreads rapidly which is more likely to cause serious consequences, especially during the COVID-19 pandemic. On the premise of meeting data privacy and security requirements, federated learning uses multi-party heterogeneous data to further promote machine learning. This paper proposes a federal learning based COVID-19 fake news detection model with deep self-attention network (FL-FNDM). We construct a deep self-attention network for fake news detection, which combines self-attention-based pretrained model BERT and deep convolutional neural network to detect fake news. Moreover, the fake news detection model is learned under the framework of horizontal federated learning, aiming at protecting users' data security and privacy. The experimental results demonstrate that the proposed model can improve the performance of fake news detection on the COVID-19 dataset, which can achieve almost the same effect of sharing data without leaking user data. © 2021 IEEE.

Prediction of the Impact of Covid-19 Vaccine on Public Health Using Twitter

Social media is a source of big data. Media like Twitter and Facebook has been used for collecting and analyzing user data for different purposes. The data can be used to analyze people opinions towards certain topics and incidents by applying sentiment analysis and then certain useful insights can be drawn from the analyzed data. During the current time of Covid-19, people have been sharing information regarding Covid-19 statistics, vaccines, and discussing the effects of the vaccine concerning public health. The purpose of this study is to analyze tweet data regarding the Covid-19 vaccine by applying sentiment analysis and predicting the impact of the vaccine on public health. Also, the tweets are analyzed for hidden topics by applying Topic Modelling using Latent Dirichlet Allocation (LDA) and Latent Semantic Analysis (LSA). The source of data for this study is Twitter API. The coding and data analysis is done using Python programming language in the Spyder (Scientific Python Development Environment) that is an integrated development environment for scientific programming, testing, and data analysis. The results of the study indicate a greater positive sentiment reflecting a healthy public discussion about the Covid-19 vaccine, information, awareness, and public acceptance. With these results, a positive impact of the Covid-19 vaccine on public health is predicted. The results of topic modeling discovered 10 hidden topics from the tweet dataset. © 2022. All Rights Reserved.

Emotion Analysis of Covid Tweets using FastText Supervised Classifier model

The world has come to a standstill due to the Covid-19 Pandemic. Managing mental and psychological health remains as important as physical health in such situations. With work from home becoming the new normal, social media has been an integral part of sharing user sentiments through tweets during these tough times which can later be used to interpret the emotions behind those tweets. Our system of study aims to perform emotion analysis by fetching user data from twitter, and analysing those tweets to understand user sentiments over a period of time. This is achieved by using Facebook’s Fasttext for text classification which classifies tweets into emotions namely Anger, Relief, Boredom, Happiness, Hate, Fun, Love, Surprise, Worry, Enthusiasm, Sadness and Empty in the most fastest and efficient way as compared to other algorithms. FastText is an NLP Library generally used for text representations and classifications. The classified sentiments over a period of time gives better understanding of people’s mental health and how the sentiments have changed overtime. © 2021 IEEE.