← Go back

Knowledge Graph Summarization (KGS)

Project Group Master

As knowledge graphs grow larger and more complex, there's a rising interest in knowledge graph summarization, a task designed to help users make sense of vast and interconnected data. Imagine you’re exploring a massive web of facts, and you want to quickly understand a specific entity, like Albert Einstein or Berlin. Entity summarization selects the most relevant and informative pieces of information (triples) about that entity, providing a concise and meaningful summary. This task is crucial in making knowledge graphs more accessible, interpretable, and useful for a wide range of applications, from search engines and recommender systems to AI assistants and data analysis platforms.

Project Goal

Our project group is excited to tackle this challenge and improve the quality of entity summaries by focusing on three important aspects:

  • Informativeness: ensuring the summary includes essential facts,
  • Diversity: capturing different facets of the entity, and
  • Frequency and Centrality: identifying important triples based on how frequently their elements appear and how well-connected they are within the graph.

To achieve this, we are exploring several cutting-edge techniques:

  • Enhancing unsupervised learning approaches that don’t rely heavily on labeled data,
  • Applying deep learning with embeddings to better capture semantic meaning, and
  • Leveraging large language models (LLMs) to generate summaries that are both accurate and human-readable.

Through this project group, students will gain hands-on experience with the full pipeline of entity summarization in knowledge graphs, from concept understanding and dataset exploration to model implementation and potential improvements.

For more information, check out the slides: KGSUMM_PG_WiSe_25.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

L.079.070506 Project Group: Knowledge Graph Summarization (in English)