About this Video

• Video Title: Linked Open Literature Review using the Neuro symbolic Open Research Knowledge Graph
• Channel: ORKG
• Speakers: Cameron
• Duration: 00:27:37

Introduction

This video demonstrates how to utilize the ORKG (Open Research Knowledge Graph) platform to create a linked open literature review. The speaker showcases the platform's features, including automated metadata extraction, knowledge graph construction, comparison table generation, and version control for literature reviews. The focus is on a case study involving food information engineering.

Key Takeaways

• Semantic Review (Review 3.0): The video introduces a semantic review approach, moving beyond static web page-like reviews (Review 1.0) to a system that identifies key insights from scientific papers and organizes them using a knowledge graph. This facilitates machine processability and human understanding.
• Neuro-symbolic Knowledge Graph: ORKG uses a neuro-symbolic knowledge graph combining connectionist (e.g., CNN, ChatGPT) and symbolic models to extract information, provide recommendations, and link to external resources like Wikidata.
• Automated and Assisted Workflow: The platform offers varying degrees of automation, from manual paper addition to fully automated metadata extraction and information extraction assisted by AI (e.g., ChatGPT). Human validation is also integrated.
• Comparison Tables and Templates: ORKG allows users to create comparison tables for different research contributions or data sources and provides templates to standardize data entry and organization.
• Version Control and Smart Review: The platform enables version control for literature reviews, allowing researchers to track changes, compare different versions, and easily view differences between revisions.

The transcript states that the platform uses the Resource Description Framework (RDF) to organize knowledge and link to external resources such as Wikidata and dbpedia. However, it does not detail how the RDF is specifically implemented or used in the knowledge graph structure. The explanation remains at a high level, describing its function as a tool for making the knowledge understandable by computers and for linking to external resources.