An empirical meta-analysis of the life sciences linked open data on the web

Area of Science:

• Biomedical Informatics
• Data Science
• Semantic Web Technologies

Background:

• The biomedical community has numerous open data sources.
• Researchers face significant challenges in discovering, querying, and integrating heterogeneous data.
• Semantic Web and linked data technologies offer potential solutions.

Purpose of the Study:

• To evaluate the semantic heterogeneity within the Life Sciences Linked Open Data (LSLOD) cloud.
• To create an LSLOD schema graph from over 80 biomedical linked open data sources.
• To identify challenges in data integration within the LSLOD cloud.

Main Methods:

• Extraction of schemas from more than 80 biomedical linked open data sources.
• Construction of an LSLOD schema graph.
• Empirical meta-analysis to assess semantic heterogeneity.

Main Results:

• Several LSLOD sources function as isolated data silos.
• Schemas are often unpublished, with limited reuse or mapping to other sources.
• Some schema elements are not conducive to biomedical data integration.

Conclusions:

• Significant semantic heterogeneity exists across the LSLOD cloud.
• The LSLOD schema graph can assist researchers in data integration.
• Addressing schema issues is crucial for effective biomedical data linkage and querying.