Projects

Current research and selected earlier work. Updated October 2026.

Explore projects on my research website ↗

Graph learning & system design

Scalable graph sampling & sparsification

Published & ongoing

From attribute-guided sampling to learned sparsifiers, support graphs, and edge-disjoint decomposition: develop sparse graph representations that preserve useful structure while reducing GNN computation and memory.

SGS-GNN ↗ · Scaffold ↗ · EDiS ↗ · AGS-GNN ↗ · Code ↗ · Code ↗ · Code ↗

GNN unification & architecture generation

Preprint & ongoing

Use a common layer equation to compare graph neural network families and explore structurally consistent architecture generation through propagation, message passing, aggregation, and transformation.

Unifying Graph Neural Networks Through a Common Layer Equation ↗

Joint hardware/software partitioning & scheduling

HPEC 2026 · accepted

Develop differentiable GNN and reinforcement learning methods for coupled partitioning and scheduling decisions in system design, accounting for resource, area, and communication constraints.

Diff-GNN ↗ · Joint Optimization of HW/SW Partitioning and Scheduling with Deep Reinforcement Learning ↗

Graph-based circuit equivalence checking

Ongoing

Investigate GNN-based subcircuit identification, gate embeddings, and graph matching to support circuit equivalence verification. Current research builds on graph representations of circuit structure.

Graph sparsification methods & benchmarks

Ongoing

Develop a comprehensive textbook and benchmarking framework spanning graph sparsification, rewiring, augmentation, theoretical guarantees, and applications. Compare methods across graph types and downstream learning tasks.

Related implementations ↗

Document intelligence & agentic AI

NEPATEC 2.0

Public dataset

Contributed to a standardized corpus of NEPA documents and metadata for research in environmental permitting and document intelligence. The public dataset connects project, process, document, and page-level information.

NEPATEC v2.0 ↗ · Hugging Face dataset ↗ · Technical report ↗

NEPATEC 2.5: multimodal metadata extraction

Ongoing

Develop multimodal LLM methods to extract metadata from heterogeneous NEPA documents, geolocate projects from map images, and generate project summaries. This is ongoing work described in the October 2026 CV.

PermitCE: AI-assisted document search & chat

Ongoing

Build a retrieval-augmented system for NEPA categorical exclusion documents, supporting semantic search, recommendations, summarization, and interactive question answering.

Cybersecurity & applied machine learning

Vulnerability mapping & threat intelligence

Published & ongoing

Connect CVEs with weakness classes and attack patterns using language models, and integrate security repositories into knowledge graphs for multi-hop analysis. Study scalability on modern AI accelerators.

V2W-BERT ↗ · Towards Automatic Mapping of Vulnerabilities to Attack Patterns Using Large Language Models ↗ · VWC-BERT ↗ · ATT&CK to CVE ↗ · Generative AI for Advanced Cyber Defense ↗ · Code ↗ · Code ↗

Composite metric for cyber risk prioritization

Ongoing

Develop a unified vulnerability prioritization metric that combines CVSS, CWSS, EPSS, and LEV to bring severity, weakness context, exploit likelihood, and exploitation evidence into a common assessment.

Predictive analytics for russet potato selection

Collaborative research

Apply machine learning, feature selection, and predictive modeling to phenotypic data to support decision-making in potato breeding.

Machine learning for the USEPA SPECIATE database

Collaborative research

Apply clustering to chemical source profiles to identify similarities in species distributions and support environmental data exploration.

Earlier work

Multi-objective & many-objective optimization

Published

Study multi-objective and many-objective evolutionary optimization, including adaptive fuzzy dominance, reference-point guidance, and the balance between Pareto convergence and solution diversity. Apply these methods to cluster performance and scheduling with competing objectives.

Evolutionary algorithm using adaptive fuzzy dominance and reference point for many-objective optimization ↗ · Exploiting a Synergy between Greedy Approach and NSGA for Scheduling in Computing Clusters ↗ · Many-objective performance enhancement in computing clusters ↗ · Code ↗

Exploiting a Synergy between Greedy Approach and NSGA for Scheduling in Computing Clusters

A. S. M. Rizvi, Tarik Reza Toha, Siddhartha Shankar Das, Sriram Chellappan, and A. B. M. Alim Al Islam.

Published · NSysS 2018

Paper ↗

Many-objective performance enhancement in computing clusters

A. S. M. Rizvi, Tarik Reza Toha, Siddhartha Shankar Das, Sriram Chellappan, and A. B. M. Alim Al Islam.

Published · IEEE IPCCC 2017

Paper ↗

A Many-objective Evolutionary Approach Using Fuzzy Dominance With Bidirectional Bias

Siddhartha Shankar Das, Monirul Islam, Naheed Anjum Arafat

[pdf][code]

Submitted as Undergraduate Thesis

A Grid Based Many-Objective Evolutionary Algorithm using Fuzzy Fitness

Siddhartha Shankar Das, Naheed Anjum Arafat, Monirul Islam

[pdf][code]

Best poster award: 3rd in 2nd undergraduate thesis poster presentation, 2014 (out of 53 submissions)

Implementation of document image Segmentation algorithm for handwritten OCR

Siddhartha Shankar Das

[presentation][code]

Advanced Image Processing Course, 2016

A short survey on Near-Duplicate Video detection and retrievalSiddhartha Shankar Das

[presentation][pdf]

High Dimensional Data management Course, 2016

Analysis of Advertisement fraudulent activities (click fraud, impression fraud etc)

based on our implementation of a large-scale Advertisement service system, that loads millions of ads daily into different news portal of Bangladesh (prothom alo, bdnews, bangla tribune, bikroy )

[code] Siddhartha Shankar Das, Azad Salam

[report]

scLogNet: Supply Chain Logistics Network

Earlier software development under the UGC-HEQEP project for Bangladesh’s food sector. This prototype combined product price, supply, and demand information with road, river, and rail network data to explore transportation options.

Diagnosis of cancer using Decision Tree learning and AdaBoost ensemble classifier, Text classification using K-Nearest Neighbor and Naive Bayes algorithm. [code]

ANN (Artificial Neural network for classification), CYK algorithm for context-free grammars, Correlation in image matching, Sequential clustering algorithms. [codes]

Pattern recognition based robot movement: E-puck robot is programmed identify and learn symbols through video image processing and moves accordingly [codes]

8-bit Microprocessor: A simple processor, capable of executing a set of 28 instructions, was designed and implemented. [proteus file]

Statistical simulation of a traffic system, Modelling and simulation of M/M/s queueing system with jokey.

python implementation

Web Development

University Course Management: A web-based system for managing courses online and a communication medium for teacher and students. [codes]

Online Course Management System 

An Oracle-based web application that manages courses of a school and university. [code]

Implementation of C Compiler

A simple ANSI C Compiler was designed and implemented.[code][python assembler]

Java IDE for developing java based application

SDJeditor: An IDE for java. Here programmer can write and execute java based application. Shows compiling errors, provides syntax highlighting. Can make executable JAR file from classes. [codes]