ORKG ASK Transforms Research

Searching for relevant scientific literature can be a huge task for researchers. With over 100 million articles available, finding what you need can be difficult. ORKG ASK, developed by the TIB – Leibniz Information Centre for Science and Technology and the L3S Research Centre, offers a solution. This innovative search and exploration system uses large language models (LLMs) alongside a structured knowledge graph to provide researchers with precise answers to their scientific questions.

Three building blocks for better research results

ORKG ASK uses a multi-level approach to make the research process more efficient:

• Semantic Search: Users can enter their questions in natural language and receive a curated selection of highly relevant articles.
• Large Language Models (LLMs): Open-source AI models extract detailed answers and summaries from these articles to provide direct and accurate insights.
• Knowledge graphs: The Open Research Knowledge Graph (ORKG) organises scientific information, links interdisciplinary findings, and improves knowledge accessibility.

Why is ORKG ASK advancing research

The system offers researchers many advantages:

• Comprehensive access to a database containing over 70 million publications from various disciplines.
• Fast and precise: The Retrieval Augmented Generation (RAG) method provides targeted answers to complex research questions.
• Interdisciplinary perspectives: ORKG ASK connects knowledge across disciplinary boundaries.
• User-friendly: Functions such as literature management, filter options, and export functions simplify everyday tasks.

ORKG ASK offers a new approach to scientific work

ORKG ASK goes beyond conventional search systems. It actively supports scientists by finding and contextualising relevant literature. The system also promotes transparency: AI-generated content clearly shows where the information comes from and how it was processed. As an open-source, open-data project, all data is freely available.

The TIB and L3S research team intend to continue developing ORKG ASK. Planned enhancements include multilingual functions, domain-specific data models and personalised search options, as well as closer integration into the research process, from idea generation to publication.

ORKG ASK is supported by a major initiative involving TIB, L3S, the European Research Council, the National Research Data Infrastructure (NFDI) and the Leibniz Association. The initiative’s shared objective is to enhance scientific searches using cutting-edge AI technology and support researchers in achieving the right findings more quickly.