Knowledge Graphs: Construction, Reasoning, and Retrieval-Augmented Generation

Project Group Master

A Knowledge Graph (KG) is a structured data model that represents real-world entities, such as people, places, or proteins, and the relationships among them using Resource Description Framework (RDF) triples in the form of (Subject, Predicate, Object). For example: (Python, used_for, Machine Learning). Unlike traditional databases that primarily store data, Knowledge Graphs establish a connected web of contextual information, enabling advanced artificial intelligence applications including semantic search, intelligent recommendation systems, and complex question answering. Contemporary KG systems are evolving from static, manually curated data entry toward dynamic, automated ecosystems based on three foundational pillars: 1. Construction: Transforming Unstructured Text into Structured Knowledge Graphs 2. Reasoning: Enabling the Knowledge Graph to Infer and Integrate Relationships 3. KG-RAG: Providing Generative AI with a Structured Factual Reference Framework

Project Objectives

Our project group aims to address the fundamental challenges of Knowledge Graph construction, reasoning, and retrieval-augmented generation throughout the entire research lifecycle, including literature review, identification of research gaps, solution formulation, experimental execution, and results analysis.

To achieve these objectives, we are investigating several advanced techniques, including:

Utilizing Large Language Models (LLMs) to construct Knowledge Graphs from unstructured text;
Applying machine learning and natural language processing techniques to enhance Knowledge Graph reasoning;
Investigating KG-RAG architectures, such as GraphRAG, in conjunction with other retrieval-augmented methodologies.

Through participation in this project group, students will acquire practical experience across the entire pipeline of Knowledge Graph construction, reasoning, and retrieval-augmented generation, encompassing conceptual understanding, dataset exploration, model implementation, and system optimization.

For more information, check out the slides: KGCR-RAG_PG_WiSe_26.pdf

FAQs

Q: What is the selection process for this project?
A: Candidates will need to submit an assignment and undergo an interview as part of the selection process.

Q: Is there a seminar connected to this PG?
A: No.

Q: What are the prerequisites for this PG?
A: The ideal candidate should possess foundational knowledge in NLP and ML, along with strong programming skills in Python and shell scripting. Additionally, proficiency in Linux is essential. The ability to learn quickly and adapt to new technologies and methodologies is also critical as the PG domain is expected to have steep learning curve.

In case you have further questions, feel free to contact Asep Fajar Firmansyah.

Course in PAUL

Project Group: Knowledge Graphs: Construction, Reasoning, and Retrieval-Augmented Generation (in English)