Research

Research Interests

• Social Network Analysis and Mining
• Distributed Systems
• Mobile Computing
• Natural Language Processing

Grants

• PRTS, ACI, PI, 33K, 2024
• College of Computing, GVSU, Seed Grant, PI, 20K, 2024
• ESC Grant, PI, 5K, 2023
• GVSU University Counselling, PI, 7K, 2023
• MI DNR & Adopt a Hemlock, PI, 32K, 2023
• Stenger & Stenger, PI, 53K, 2023
• Array Of Engineers Grant, PI, 38K, 2023
• Special Project GA, GVSU, 9K, 2022-2023
• Array Of Engineers Gift, PI, 45K, 2022
• P&G Gift, Co-PI, 50K, 2022
• Special Project GA, GVSU, 9K, 2021-2022
• CSCE Grant, GVSU, 3K, 2021
• P&G Gift, Co-PI, 75K, 2021

Current Research Projects

GoFundMe (GFM) Data Collection and Analysis

This project involves the following. First, collect GFM data for future analysis. Second, use NLP and ML techniques to predict the category of a fundraiser (emergency, community, education) based on the description of the fundraiser. Three, understand how the fundraiser behavior is different across different categories. Fourth, can we predict the success probability of a given fundraiser given the initial donation time series of the samaritans.

Oak Wilt Detection in MI State Forests

Partnering with Adopt a Hemlock and MI DNR to leverage computer vision and UAVs to facilitate early detection of oak wilt in Michigan

HITL-NLP Powered approach to visualize Gene Pathway Research

Collaboration with Dr. Guenter Tusch to develop an nteractive dashboard for research into gene pathways.

Developing honeypot when cyberbullying takes place

Collaboration with Dr.Sara Sutton to develop a system where bullying perpetrators are lured into a honeypot where the system analyzes bullying behaviors accordingly, mimicing victim and upstander roles accordingly.

NLP and HITL towards automated software test generation

Collaboration Array of Engineers to develop a system that can generate safety critical software tests from requirements.

Several stealth startup projects

Working with multiple students on several stealth startup ideas.

Current Students

• SM Azizul Hakim (Working on Array of Engineers project)
• Usman Tahir Qureshi (Working on P&G project)
• Mohammad Shafiq (Working on Cyberbullying Research)
• Grant Alphenaar (Current ACI Research Scientist)
• Muttaki Islam Bismoy (Working on Oak Wilt project with Michigan DNR)
• Nazmus Sakib (Working on Stenger and Stenger project)
• Aliah Lloyd (Working on ESC Grant project on mental health app)
• Aastha Thapa (Working on GV Mental Health App)

Past Students

• Abu Naweem Khan (Now at Dematic)
• Debit Paudel (Now at Mitsubishi)
• Griffin Going (Now at RankOne Computing)
• Esteban Echeverri Jaramillo (Now pursuing PhD at MSU)
• Alvaro Ardila Perez (Now at Facebook)

Masters Thesis

• Grant Alphenaar : Predicting Course Performance on a Massive Open Online Course Platform: A Natural Language Processing Approach

Previous Projects

YouBrush: Leveraging Edge-Based Machine Learning In Oral Care

Abstract: A disconnect is frequent regarding the length of the time a person claims to have brushed their teeth and the actual duration; the recommended brushing duration is 2 minutes. This paper seeks to bridge this particular disconnect. We introduce YouBrush, a low-latency, low-friction, and responsive mobile application to improve oral care regimens in users. YouBrush is an IOS mobile application that democratizes features previously available only to intelligent toothbrush users by in corporating a highly accurate deep learning brushing detection model developed by Apple’s createML on the device. The machine learning model, running on the edge, allows for a low-latency, highly responsive scripted-coaching brushing experience for the user. Moreover, we craft in-app gamification techniques to further user interaction, stickiness, and oral care adherence.

BullyAlert: an Adaptive Cyberbullying Detection Mobile Application for Parents

Abstract: BullyAlert is an android mobile application that has been developed for the parents to monitor online social network activities of their kids and get notifications when a potential Cyberbullying instance takes place. This application implementents adaptive classifier mechanisms to make sure the notifications that an individual parent receive are calibrated according to their tolerance level.

Understanding LGTBQA+ Cyberbullying Behavior in Online Twitter Communities

In this project, we are using Data Mining, Machine Learning , Natural Language Processing and Community finding techniquesto understand how cyberbullying languages directed to LGBTQA+ communities evolved through the years across online communities in Twitter.

Rate My Professors (RMP) Data Collection and Analysis

This project involves the following. First, collect RMP data for future analysis. Second, understand how student expectations,reviews vary across universities of different types (R1,R2,teaching), location (east coast, west coast etc) and departments (computer science, social science etc). We are then interested in predicting the quality of a professor using the reviews and other metadata collected from the website.

Multi-modal Fusion for Flasher Detection in a Mobile Video Chat Application

Using multi-modal mobile sensor data and temporal data to substantially improve the accuracy of the fusion classifier, compensating for the loss in accuracy due to the weaker correlation between facial absence and flasher behavior in MVChat, a mobile video chat application.

Investigation of Cyberbullying Behavior in Ask.fm, Vine and Instagram

Detailed investigation of cyberbullying behavior in online social networks by collecting data, labeling the data, perform analysis of the labeling data and then building an accurate classifier.

Scalable and Timely Detection of Cyberbullying in Online Social Networks

Leveraging dynamic priority scheduler and incremental classifier computation, we were able to build a system that is five times more scalable resource-wise and seven times more responsive.