Current Project Openings (Updated on February 19, 2025)
DREAM Committee: Nazli Ikizler Cinbis and Erkut Erdem
Molecular Intelligence for Novel Discovery (MIND): A Universal Multi-Modal Foundation Model for Life’s Molecules
Name of the faculty mentor(s): Tunca Doğan (tuncadogan@gmail.com)
Project Description: The aim of this project is to develop a universal foundation model for molecular representation learning, capable of handling all molecules related to life, including biomolecules such as DNA, RNA, proteins, and small molecules such as drug-like compounds and metabolites. The proposed foundation model will employ cutting-edge deep learning techniques to encode atomic and molecular-level information and will serve as the backbone for downstream tasks. The foundation model will adopt a multi-modal learning approach to integrate diverse data types, including sequences (e.g., nucleotide and amino acid), structures (e.g., atomic 3D coordinates), interactions (e.g., protein-protein or protein-ligand networks), textual annotations, and optionally imaging data (e.g., immunohistochemistry). We will explore advanced transformer-based architectures as the primary modelling approach, including encoder-decoder models and dense attention mechanisms. Alternatives such as graph neural networks (GNNs) and diffusion models will be considered to better capture relational and spatial information inherent in molecular systems. This will enable the model to operate at the atomic level, using atoms and bonds as the fundamental building blocks to provide a unified representation for all molecules. Following pretraining, the model will be fine-tuned for generative tasks such as designing novel proteins with specified functions and drug candidates targeting biomolecules of interest. Discriminative tasks, such as protein function prediction and molecular property prediction, will leverage fine-tuned classifiers or regressors. Potential applications in biotechnology include enzyme design, metabolite synthesis, material science, and drug discovery.
Number of students: 3 students
Requirements and Additional Info: Experience: Machine learning and deep learning, high motivation and interest: generative AI, LLMs, bioinformatics
Gait analysis for measuring the performance of rehabilitation
Name of the faculty mentor(s): Pinar Duygulu Sahin (pinar@cs.hacettepe.edu.tr)
Project Description: Tracking the movements of patients in their home environment with a cell phone will provide valuable benefits in understanding the effect of rehabilitation treatments. In this project, it is required to develop a system that can track the movements of patients and give feedback during walking. You will be mostly responsible from the mobile application that can provide the on time feedback.
Number of students: 2 students
Requirements and Additional Info: Preferably, understanding of machine learning and computer vision terms.
Understanding Radiology Images with Large Multimodal Models (LLMs)
Name of the faculty mentor(s): Erkut Erdem (erkut@cs.hacettepe.edu.tr)
Project Description: In this project, the students will be working on developing large multimodal models to understand radiology images by using vision and NLP techniques. The participants will particularly investigate different approaches to combine vision models with LLMs, and train and evaluate these approaches on datasets consisting of medical data that contain radiology images and text. By the end of the project, the student will become familiar with the recent multimodal generative models.
Number of students: 3 students
Requirements and Additional Info: Candidates should have strong programming skills and solid machine learning background. Prior experience with LLMs and vision models are suggested. This project will be carried out in collaboration with Dr. Aykut Erdem of Koc University, Istanbul.
Traversable Path Finding by Using Past Experience
Name of the faculty mentor(s): Ozgur Erkent (ozgurerkent@hacettepe.edu.tr)
Project Description: In this project, you are expected to write an algorithm for a mobile robot that will learn to find the path to reach a destination. The robot will explore the environment and find a path to the goal. The robot will use its sensors such as cameras, lidars, imus, gnss, etc and a GPU based embedded system. First, you will develop algorithms on simulation, then you will integrate it on a platform.
Number of students: 2 students
Requirements and Additional Info: Interest in robotics.
Benchmarking User Interfaces of OSS Quality Evaluation Tools
Name of the faculty mentor(s): Ayca Kolukisa Tarhan (atarhan@cs.hacettepe.edu.tr)
Project Description: The user interfaces (UI) of OSS quality evaluation tools play a crucial role in how effectively users can assess and interpret software quality metrics. However, the usability, clarity, and effectiveness of these interfaces vary significantly. In this project, the student will conduct a comparative analysis of the UI/UX designs of existing OSS quality evaluation tools. The study will assess factors such as navigation efficiency, information presentation, user experience, and accessibility. Design of the study will be provided by the mentor. The findings of the study will be used to identify best practices, highlight usability gaps, and provide recommendations for improving OSS quality evaluation tool interfaces. The results will contribute to both software engineering and human-computer interaction research.
Number of students: 3 students
Requirements and Additional Info: Students should have an interest in software usability analysis and open-source software. Experience with UI/UX evaluation methods, usability testing, or empirical research in human-computer interaction is a plus but not required. Students will not be required to work closely together.
Assessing the Impact of Generative AI on Software Engineering Documentation and Learning
Name of the faculty mentor(s): Ayca Kolukisa Tarhan (atarhan@cs.hacettepe.edu.tr), Ebru Gokalp-Aydin, Tugba Erdogan
Project Description: The landscape of software engineering (SE) education is currently experiencing a significant evolution, driven by the integration of generative AI tools into industry practices [1]. These tools have the potential to markedly enhance both efficiency and quality in software development processes [2]. However, their impact on pedagogical approaches and documentation practices within academic settings necessitates further empirical investigation [3]. As referenced in recent literature [4][5], there exists an urgent imperative for educational institutions to recalibrate their curricula and instructional methodologies to encompass generative AI concepts, tools, and applications. In terms of curriculum development, it is critical to identify the specific generative AI-related competencies—such as prompt engineering, AIOps, and AI-human interaction—that should be woven into core SE subjects, including software requirements, software architecture, software design, software construction, software testing, and SE management.
This project aims to explore the effect of generative AI tools on documentation quality and student learning outcomes within undergraduate SE courses, with the pre-requisite to obey to the ACM’s principles on using these tools [6]. The broader objective is to generate actionable insights that can effectively bridge the gap between traditional educational paradigms and AI-enhanced practices in the software development lifecycle.
The research will involve an evaluation of industry-standard generative AI tools that assist in software documentation and implementation. The selection criteria for these tools will include their proven effectiveness in real-world industry contexts and their capacity to enhance software documentation practices. The tools examined in this study encompass the following categories:
Large Language Models:
• ChatGPT – Natural language documentation and requirements elicitation.
• Claude AI – Generation of detailed technical documentation.
• Google Bard – Collaborative documentation and stakeholder communication.
Development Tools:
• GitHub Copilot – Code documentation and implementation assistance.
• TabNine – Code completion and API documentation generation.
Architecture and Design Tools:
• ArchiMate AI – Architectural diagram generation and documentation.
• Enterprise Architect AI – Design pattern recognition and optimization.
Testing Tools:
• DeepCode – Automated test case generation and bug detection.
• Selenium AI – Test script documentation and execution.
• Testim.io – AI-driven test coverage analysis.
Management Tools:
• Jira AI – AI-assisted project scheduling and risk assessment.
• Monday.com AI – Resource allocation and task tracking.
The participants in this study will consist of undergraduate students enrolled in SE laboratory courses, all of whom possess foundational knowledge in programming and software design.
Methodology: A mixed-methods approach will be employed, combining quantitative and qualitative analysis. Specific guidance for executing these methods will be provided by the mentors during the project.
This research addresses the imperative need to elucidate how generative AI tools can be effectively integrated into SE education. The findings of this study aim to provide unique insights into the educational and practical applications of these tools, thereby contributing to the advancement of software engineering pedagogy.
References
[1] A. Neyem, L. A. González, M. Mendoza, J. P. S. Alcocer, L. Centellas and C. Paredes, “Toward an AI Knowledge Assistant for Context-Aware Learning Experiences in Software Capstone Project Development," in IEEE Transactions on Learning Technologies, vol. 17, pp. 1599-1614, 2024, https://doi.org/10.1109/TLT.2024.3396735.
[2] Valerio Terragni, Annie Vella, Partha Roop, and Kelly Blincoe. 2025. “The Future of AI-Driven Software Engineering”. ACM Trans. Softw. Eng. Methodol. Just Accepted (January 2025). https://doi.org/10.1145/3715003.
[3] Nguyen, K.V. “The Use of Generative AI Tools in Higher Education: Ethical and Pedagogical Principles”. J Acad Ethics (2025). https://doi.org/10.1007/s10805-025-09607-1
[4] Junda He, Christoph Treude, and David Lo. 2025. “LLM-Based Multi-Agent Systems for Software Engineering: Literature Review, Vision and the Road Ahead”. ACM Trans. Softw. Eng. Methodol. Just Accepted (January 2025). https://doi.org/10.1145/3712003.
[5] Nguyen-Duc, A., Cabrero-Daniel, B., Przybylek, A., Arora, C., Khanna, D., Herda, T., ... & Abrahamsson, P. “Generative Artificial Intelligence for Software Engineering--A Research Agenda”. arXiv preprint arXiv:2310.18648.
[6] ACM Technology Policy Council. “Principles for the development, deployment, and use of generative AI technologies”. June 27, 2023. (Last access: Feb 17, 2025) https://www.acm.org/binaries/content/assets/public-policy/ustpc-approved-generative-ai-principles
Number of students: 3 students
Requirements and Additional Info: Preferably, the student(s) should have been enrolled in (this semester or before) BBM384-SE Laboratory course. If not, s/he should be at least at his/her 4th semester with a GPA not less than 3.0.
Deep inferring miRNA expression from scRNA-seq data
Name of the faculty mentor(s): Gulden Olgun (guldenolgun@cs.hacettepe.edu.tr)
Project Description: miRNAs are small non-coding RNAs that typically bind to the 3′ untranslated regions (UTR) of their target mRNAs. They play critical roles in most fundamental cellular processes and are implicated in several diseases, including cancer. Single-cell RNA sequencing (scRNA-seq) technologies have advanced our understanding of molecular mechanisms at the single-cell level. However, these technologies are yet to benefit the field of miRNAs because of the current scRNA-seq protocol. Thus, in this project, we aim to infer miRNA activity from gene expression data using deep learning to overcome this technological challenge.
Number of students: 3 students
Requirements and Additional Info: Preferably, the student(s) should have been enrolled in (this semester or before) BBM384-SE Laboratory course. If not, s/he should be at least at his/her 4th semester with a GPA not less than 3.0.
Blockchain Simulator
Name of the faculty mentor(s): Adnan Ozsoy (adnanozsoy@gmail.com)
Project Description: In blockchain research, there is a growing need for a simple yet effective blockchain simulator that allows researchers, educators, and developers to quickly experiment with blockchain mechanisms without the complexity of full-scale implementations. Many existing blockchain frameworks are too intricate, requiring extensive setup and deep protocol knowledge, making rapid prototyping and conceptual testing difficult. A lightweight, user-friendly simulator would provide an accessible platform for trying out different consensus algorithms, transaction models, network behaviors, and security vulnerabilities in a controlled environment. Such a tool would accelerate blockchain research and education by enabling easy parameter adjustments, real-time visualizations, and rapid feedback loops, allowing users to focus on innovation rather than infrastructure. In this project we propose The Blockchain Simulator which is a comprehensive framework designed for modeling, analyzing, and simulating blockchain networks and cryptocurrency ecosystems. The simulator provides an interactive and extensible platform for researchers, developers, and students to experiment with blockchain protocols, consensus mechanisms, transaction models, and network behaviors under various conditions.
Number of students: 2 students
Requirements and Additional Info: None.
Consensus Algorithm Repository for Blockchain Research
Name of the faculty mentor(s): Adnan Ozsoy (adnanozsoy@gmail.com)
Project Description: The Consensus Algorithm Repository is a structured database and implementation hub for blockchain consensus mechanisms. This project aims to either develop or systematically collect implementations of various consensus algorithms, providing a comprehensive, standardized, and easily accessible resource for researchers, developers, and educators. By compiling different consensus models, the repository will serve as a benchmarking and experimentation platform, enabling users to analyze, compare, and integrate various consensus protocols into blockchain networks.
Number of students: 2 students
Requirements and Additional Info: None.