In biomedical literature, overviews of systematic reviews (OoSRs) have recently become a popular approach of evidence synthesis where the unit of synthesis is the systematic review (SR) [1,2]. The outcome data from SRs included in an OoSRs can be presented exactly as they are reported in the SRs (qualitative synthesis) or they can be re-analyzed using meta-analysis (quantitative synthesis). OoSRs can provide valuable information to support decision-making by healthcare professionals and to guide integration of research evidence in policy-making within a short timeframe [3].
However, traditional methods often face challenges in capturing the complex relationships and nuances within these high-level synthesis study designs and representation with knowledge graphs may be helpful. A knowledge graph represents a network of entities and illustrates the relationship between them [4,5]. It is made up of three main components: nodes, edges, and labels. This information is usually stored in a graph database and visualized as a graph structure, prompting the term knowledge “graph.”
The primary objective of this project is to develop a methodology for modeling healthcare overviews as scholarly knowledge graphs using Cypher [6], which is specifically designed for expressing patterns in graph data and performing graph operations, and advanced Natural Language Processing (NLP) techniques [7] for biomedical text mining. Our specific goals include:
Objective 1: Develop a comprehensive knowledge graph framework that accurately represents the structure and associations within healthcare overviews.
Objective 2: Utilize Cypher for efficient querying and manipulation of the knowledge graph to extract meaningful insights.
Objective 3: Apply advanced NLP techniques to extract, process, and analyze information from healthcare studies.
Objective 4: Design a user-friendly interface for researchers and clinicians to interact with the knowledge graph for facilitating exploration and knowledge discovery, and informed decision-making.