Open Research Knowledge Graph

Science is facing a challenge. In the ever-growing flood of publications, researchers find it increasingly difficult to keep an overview over relevant articles. Current search systems offer a search for articles, but not a search for the knowledge inside these articles.

This is where the Open Research Knowledge Graph (ORKG) and the AI tool ORKG Ask come into play.

With ORKG Ask, researchers can search close to 80 million publications for answers to their research questions. A semantic search determines the most relevant articles and an Large Language Model (LLM) extracts the information. In addition to a detailed overview over the knowledge inside the articles, researchers are provided with a short summary of the five most relevant answers. Researchers can then further filter the results, e.g. with regards to publication time, citation count or author. Interesting articles and questions can be bookmarked and a preselected literature list can be uploaded.

The ORKG aims to go one step further and tackle the underlying issue. By representing the knowledge inside research papers in a knowledge graph, we make them machine-actionable. This enables completely new machine assistance for finding, comparing and reusing scientific findings.

By combining these two different approaches – LLMs and Knowledge Graphs – the ORKG is becoming a lighthouse in the publication flood.