Available Projects for RET 2025

Research Experiences in Big Data and Machine/Deep Learning for OK STEM Teachers.

Research Projects Overview

Project 1 (P1): Enhancing Road Safety Through Secure Vehicular Networks

Categories: Vehicular Networks, Network Security

Overview: With over 42,000 traffic-related fatalities in the U.S. in 2022, vehicular networks offer a potential solution by providing real-time traffic information. However, their security is a critical challenge. This project focuses on combating false information attacks within vehicular networks through innovative security techniques.

Research Goals:

1. Develop techniques to detect and mitigate false information attacks.
2. Evaluate the efficiency and accuracy of these techniques. Teacher Involvement: Educators will explore vehicular network architecture, security challenges, and network simulators for hands-on learning.
3. Data Source: Public VeReMi dataset on Kaggle.

Project 2 (P2): Understanding Social Media Polarization Through Information Cascades

Categories: Data Mining, Network Science, Machine Learning

Overview: Social media fuels polarized opinions, influencing public responses to major events like COVID-19. This project maps and quantifies political bias within news propagation networks.

Research Goals:

1. Identify information networks using social media response data.
2. Characterize polarization and predict its evolution. Teacher Involvement: Hands-on training in social media analysis, multimodal data analytics, and machine learning.
3. Data Source: Reddit, Twitter, and Gab.com datasets.

Project 3 (P3): Soil Moisture Prediction Using Smartphone Images

Categories: Digital Image Processing, AI for Agriculture

Overview: Farmers lack site-specific soil moisture prediction tools. This project develops a deep learning model to estimate soil moisture levels from smartphone images.

Research Goals:

1. Create models that assess soil moisture from photographs.
2. Improve drought prediction and optimize irrigation schedules. Teacher Involvement: Hands-on experience with machine learning techniques, including object detection and image classification.
3. Data Source: Agricultural datasets from collaborating researchers.

Project 4 (P4): Mitigating Echo Chambers in Online Social Media

Categories: Natural Language Processing, Text Generation

Overview: Existing summarization tools fail to neutralize political bias in online discussions. This project develops AI-driven summaries that provide unbiased perspectives.

Research Goals:

1. Develop unsupervised models to extract unbiased narratives from online debates.
2. Analyze user perception and prior exposure to discussions. Teacher Involvement: Training in sentiment analysis, AI-based summarization, and NLP techniques.
3. Data Source: Twitter, Reddit, and Gab.com discussions.

Project 5 (P5): Parallel AI-Based Security for Vehicular Networks

Categories: Vehicular Network Security, GPU Computing, Deep Learning

Overview: This project explores the application of GPU-based parallel computing to enhance vehicular network security using machine learning techniques.

Research Goals:

1. Identify ML/DL-based solutions suitable for GPU parallelization.
2. Implement and evaluate security solutions using GPUs. Teacher Involvement: Hands-on experience in GPU-based machine learning applications and parallel computing.
3. Data Source: Generated using the SUMO simulator.

Project 6 (P6): Detecting False Information Collusion in Vehicular Networks

Categories: Vehicular Network Security, Machine Learning

Overview: Malicious vehicles can coordinate to spread false information. This project develops unsupervised ML techniques to detect and mitigate such attacks.

Research Goals:

1. Design ML models for detecting collusive false information attacks.
2. Evaluate detection accuracy and efficiency. Teacher Involvement: Training in unsupervised ML techniques and network security simulations.
3. Data Source: DSRC Vehicle Communications dataset from the UCI repository.

Project 7 (P7): ML-Based Flooding Attack Detection in IoD Systems

Categories: Security, Networking, Machine Learning

Overview: The Internet of Drones (IoD) faces security threats from malicious flooding attacks. This project aims to develop AI and blockchain-based solutions to detect and mitigate these attacks.

Research Goals:

1. Investigate flooding attack impacts on IoD.
2. Develop ML and blockchain-based detection methods. Teacher Involvement: Training in cybersecurity fundamentals and network simulations.
3. Data Source: IoD system traces.

Project 8 (P8): Automated Cyber Threat Intelligence Extraction Using XAI

Categories: Autonomous Systems, Cybersecurity, Explainable AI

Overview: Autonomous systems require real-time cybersecurity strategies. This project leverages Explainable AI (XAI) to automate cyber threat intelligence (CTI) extraction.

Research Goals:

1. Develop ML models for extracting CTI from predicted threats.
2. Structure CTI using the STIX framework. Teacher Involvement: Training in AI-based security analysis and cyber threat intelligence.
3. Data Source: IIoT attack datasets from public repositories.

Project 9 (P9): Anomaly Detection in Smart Agriculture

Categories: Smart Agriculture, AI, Anomaly Detection

Overview: IoT-driven smart agriculture is vulnerable to cyber threats. This project uses AI techniques to detect anomalies in farming data.

Research Goals:

1. Develop ML models to identify security threats in smart agriculture.
2. Utilize XAI methods to improve transparency in anomaly detection. Teacher Involvement: Hands-on experience in AI-driven security applications.
3. Data Source: IoT botnet datasets from UNSW Science repository.

Project Alignment with Oklahoma Academic Standards for CS (High School)

Oklahoma Academic Standards Computer Science