Skip to content
SHShahadat HossainAssistant Professor · DIU

Research

I study how people behave on digital platforms, combining machine learning, behavior analysis and information visualization to explain what drives their decisions.

Research areas

Four connected areas, each with its related publications.

01

Human Behavior Analysis

The core of my work is understanding human behavior on digital platforms: the factors that drive consumer decisions in Facebook-based commerce (F-commerce), how readers engage with online news, how sentiment and personality cues form around public figures on Twitter, and how users protect or expose themselves on social networks.

Guiding questions

  • Which factors most influence consumer behavior in social commerce?
  • What reading patterns characterize online news audiences?
  • How do sentiment and personality traits surface in social media discourse?

Related publications · 8

3 more
02

Data Science & Information Visualization

I use exploratory data analysis, clustering and visualization to make large, noisy socio-economic datasets legible, from the World Crime Index and the Corruption Perception Index to stock-market sectors during COVID-19 and big-data storage practice. The aim is insight that decision-makers can read at a glance.

Guiding questions

  • What structure hides in global indices once we cluster and visualize them?
  • How can big data analytics capabilities be measured and communicated?

Related publications · 7

2 more
03

Machine Learning & Deep Learning

I apply supervised and unsupervised learning (ANN, Random Forest, KNN, DBSCAN, PCA, CNN and transformer-based NLP) to behavioral and applied problems, including fake-news detection, Bangla voice assistance, crop disease detection and personalized recommender systems for music therapy.

Guiding questions

  • Can learning models reliably flag fake news and hostile text, including in Bangla?
  • How can recommender systems personalize care-oriented services?

Related publications · 10

5 more
04

Information Systems, Security & SEM

Complementing the analytical work, I study secure and practical information systems (blockchain land administration, cryptographic wireless networks, banking-network security, database choices for MIS) and use Structural Equation Modeling (SEM) to test how users perceive and adopt technology.

Guiding questions

  • What security trade-offs do users accept in everyday digital services?
  • Which latent constructs explain technology adoption in emerging economies?

Related publications · 11

6 more

Methods and techniques

  • Artificial Neural Networks
  • Random Forest
  • K-Nearest Neighbors
  • DBSCAN Clustering
  • Linear Regression
  • Principal Component Analysis
  • Convolutional Neural Networks
  • Sentiment & Polarity Analysis
  • BERT / SVM
  • Structural Equation Modeling
  • Exploratory Data Analysis
  • Recommender Systems
  • Blockchain
  • Ant-Colony Optimization

Published with

  • Springer
  • IEEE
  • ACM
  • Taylor & Francis
  • IGI Global
  • AIP

Frequent co-authors

Researchers I have published with more than once.

  • MHMd. Manzurul Hasan22 papers
  • MBMimun Barid5 papers
  • MRMd. Mahmudur Rahman4 papers
  • THTanvir Hossain3 papers
  • DGDipta Gomes2 papers
  • GHGahangir Hossain2 papers
  • KSKazi Sazzad Hosen2 papers
  • TMTanzila Mehnaz2 papers