Search Constraints

Start Over
Filtering by: Department - School of Information Technology Remove constraint Department: - School of Information Technology

Search Results

  • IoT Emergency Detection Devices

    Type:
    Article
    Description/Abstract:
    In a world where technology continues to vastly grow and improve, IoT devices have increasingly become more and more a part of people’s everyday lives. Although that is the case there is a need to understand how to better use these devices for threat detection. This paper presents early work to understand gaps in this regard using a review of previously used techniques to identify known threats to households. Through the use of smart home device clusters we seek to effectively reduce the amount of false alarms and create a more reliable resource for home residents.
    Creator/Author:
    Tresenwriter, Isaac
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Securing A Small Office Home Office Network: Integrating A Supervised Neural Network Algorithm

    Type:
    Article
    Description/Abstract:
    Small office home office networks have become a target for many threat actors, hackers and cyber attackers and hence there is an urgent need to secure the network from such attackers. Most small office home office network users do not see the need to provide enough security to their networks because they assume no one is going to hack them forgetting that the biggest threat of our small home networks today comes from the outside. The challenge of misconfiguration of routers, firewalls and default configurations in our small home networks renders the network vulnerable to attacks such as DDos , phishing attacks , virus and other network attacks hence the need to implement a detection algorithm to help identify flaws in the pattern of the small office network. It turns out that about 75% of existing approaches focused on intrusion detection in 802.11 wireless networks of a SOHO and not the entire network. These approaches do not efficiently secure the network entirely leaving the rest prone to attacks can occur with or without the internet. This paper proposes to add another layer of security to the other preventive measures in a SOHO network by designing, implementing and testing a supervised neural network algorithm to identify attacks on the small home network and also to send a notification to users to keep them informed of the activities on their network. The supervised neural network algorithm will have a dataset representing both attacks and non-attacks which will be used in the training phase. The system should be able to detect and identify the various attacks and anomalies when they occur on the network and help keep the users informed.
    Creator/Author:
    Azumah, Sylvia Worlali and Li, Chengcheng
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Implementation of a hybrid machine learning technique for network intrusion detection

    Type:
    Article
    Description/Abstract:
    The current rapid growth in the computer and internet development has ushered in numerous cybersecurity challenges which are constantly evolving with time. The current cybersecurity solutions are no longer optimal in tackling these emerging cyber threats and attacks. This paper proposes the creation of a cybersecurity dataset to be used for a hybrid machine learning (ML) approach of supervised and unsupervised learning for an effective intrusion detection system. The proposed model entails a five-stage process which starts at the setup of a simulated network environment of network attacks to generate a dataset which feeds into the data normalization stage and then to data dimension reduction stage using the principal component analysis as a feature extraction method after which the data of reduced dimension is clustered using the k-Means method to bring about a new data set with fewer features. This new dataset is afterward classified using the enhanced support vector machine (ESVM). The proposed model is expected to provide a high-quality dataset and an efficient intrusion detection system in terms of intrusion detection accuracy of 99.5%, short train time of 3seconds and a low false-positive rate of 0.4%.
    Creator/Author:
    Eichie, Maxwell
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Crime Hot Spots: Proposing a new model for appropriate identification of crime hot spots

    Type:
    Article
    Description/Abstract:
    This research focuses on two fundamental aspects of hot spot policing that have been widely neglected by previous scholarly research. These aspects include the adequate concentration of crime at a smaller geographical unit to be considered a crime hot spot, and the cost-benefit implication of focusing limited police resources on such a smaller place in an effort to prevent criminal activities. Substantial limitations in call-t- service data from police departments raise concern on the purported concentration of crime at places that warrant such strategy in the first place. We will examine data from the Cincinnati Police Department and propose guidelines on adopting a threshold when designating places as crime hot spots, using time and cost-benefit analysis as key determinants.
    Creator/Author:
    Hussein, Abdul Aziz and Ozer, Murat
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Scraping The Deep Web: A 3-Dimensional Framework For Cyber-Threat Intelligence

    Type:
    Article
    Description/Abstract:
    Cyberspace is one of the most complex systems ever built by humans. The utilization of cybertechnology resources are used ubiquitously by many, but sparsely understood by the majority of the users. In the past, cyberattacks were usually orchestrated in a random pattern of attack to lure unsuspecting targets. However, the cyber virtual environment is an ecosystem that provided a platform for an organized and sophisticated approach to launch an attack against a specific target group or organization by nefarious actors. In 2019, the average cost of cyber-attack in the US was about $1.6 million. This paper proposes a 3D framework to signal new threat alert before the actual occurrence of the threat on the surface web to alert cybersecurity experts and law enforcement agencies in preventive measures or means of mitigating the severity of damage caused by cyberattacks. The methodology combines information extracted from the deep web through a smart web crawler with socio-personal and technical indicators from twitter which is mapped with OTX (Open Threat Exchange). The OTX is an open-source cyber threat platform managed by security experts. The OTX endpoint security tool(OTX python SDK) will be used to identify a new type of cyber threats. The effectiveness of the framework will be tested using the machine learning algorithm precision-recall rate.
    Creator/Author:
    Adewale, Adewopo Victor and Gonen, Bilal
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Relative Importance of Blockchain in Health Care: Enhancing data sharing and mitigating cyber security crimes

    Type:
    Article
    Description/Abstract:
    This paper looks at the opportunities and challenges of implementing blockchain technology across the medical sector and provides a clear view which can enable blockchain for more extents. After a notable research on underlying blockchain technology which offers distributed governance, immutable audit trail, provenance of data, robustness and privacy, we contrasted blockchain innovations and identified prominent applications of it in historically decentralized healthcare sectors. As the healthcare industry faces many challenges like unauthorised data sharing, lack of data transparency, ransomware, data breaches and cyber crimes, blockchain is one of the best ways to enhance data sharing and to mitigate prominent cyber crimes. By proper designing of a decentralized and immutable blockchain network where the data is dispersed among credentialed social insurance experts guarantees that cybercriminals cannot touch single patient’s confidential data, which facilitates encryption or cryptography of personal data where no patient’s emergency data is at extreme hazard. Blockchain trust-worthy cloud is one of the most powerful and secure ways of storing high confidential data. After analysing Blockchain implementations and identifying its potential in healthcare, we conclude with several promising directions for future research.
    Creator/Author:
    Ponnakanti, Hari Priya; Ozer, Murat, and Gonen, Bilal
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • A methodology for cyberthreat ranking

    Type:
    Article
    Description/Abstract:
    National Institute of Standards and Technology (NIST) recommends that organizations perform cyber risk assessments regularly to identify security vulnerabilities and to control levels of exposure to threats. We discuss a method to customize the ranking of cyber threats based on the organization’s maturity level of implementing NIST controls and we use FAIR model’s LEF component as a measure of the severity of cyber threats. The methodology integrates NIST maturity levels to calculate the resistance strength factor and produce the LEF values for each threat. The LEF value is then used to represent the severity level of the threat to the specific organization. This hybrid risk assessment approach will help stakeholders make data-informed decisions on improving security measures and provide accurate values that represent the current security state of their organization.
    Creator/Author:
    Bakare, Adeyinka and Said, Hazem
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Application of Autoencoder Duets in Anomaly and Intrusion Detection

    Type:
    Article
    Description/Abstract:
    Signature-based intrusion detection methods report high accuracy with a low false alarm rate. However, they do not perform well when faced with new or emerging threats. This work focuses on anomaly-based data driven methods to identify potential zero-day-attacks using a specific class of neural networks known as the autoencoder.
    Creator/Author:
    Li, Chengcheng; Lee, Kijung; Mathur, Nitin, and Gonen, Bilal
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Explaining the Relationship between Internet and Democracy in Partly Free Countries Using Machine Learning Models

    Type:
    Article
    Description/Abstract:
    Previous studies have offered a variety of explanations on the relationship between democracy and the internet. Some argue that with free access to information, knowledge sharing without any constraint, and the spread of political knowledge, the internet will help change people’s political attitudes and spread democracy. Other studies found that authoritarian regimes by censoring the internet, tracking the political activist, prosecuting the dissidents, and using the internet to spread their propaganda limit the democratization. Also, some studies explored the effects of diffusion of false news through the internet and especially via social media. However, most of these studies concentrate on regions, specific states or authoritarian regimes. No study has investigated the influence of the internet in partly free countries defined by the Freedom House. Moreover, very little is known about the effects of online censorship on the development, stagnation, or decline of democracy. To fully understand the impact of the internet and online censorship on democratization in partly free countries, we must explore these relationships in these countries. Drawing upon the International Telecommunication Union, Freedom House, and World Bank reports and using machine learning methods, this study sheds new light on the effects of the internet on democratization in partly free countries. The findings suggest that internet penetration and online censorship both have a negative impact on democracy scores and the internet’s effect on democracy scores is conditioned by online censorship. Moreover, results from random forest suggest that online censorship is the most important variable followed by governance index and education on democracy scores.
    Creator/Author:
    Varlioglu, M. Said and Sagir, Mustafa
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved

  • Social Networks of Technology Caregivers and Caregivees

    Type:
    Article
    Description/Abstract:
    The paper focuses on exploring the social networks of technology caregivers and caregivees and also work on learning their preferred mode of information exchange. Responses from the participants of the study will throw light on the relationships between different efficacies (discussed in detail in the paper) that may have an impact on an individual’s decision. Participant’s responses are recorded through well constructed surveys that have been distributed around by word of mouth or specific social media platforms which will also prove if being a power user has any effect on the end result. The responses will be analyzed and the various efficacy constructs such as self efficacy, community collective efficacy will be kept in mind.
    Creator/Author:
    Kaushik, Sanjana and Elrod, Nathan J.
    Submitter:
    Jess Kropczynski
    Date Uploaded:
    05/15/2020
    Date Modified:
    05/15/2020
    Date Created:
    2020-04-14
    License:
    All rights reserved