You are here

GOClonto: an ontological clustering approach for conceptualizing PubMed abstracts

As biomedical science progresses, there is an overwhelming amount of textual knowledge being recorded in the biomedical literature.
PubMed is the most comprehensive database collecting and managing biomedical literature. To help researchers easily understand collections of PubMed abstracts, numerous clustering methods have been proposed to group similar abstracts based on their shared features.
However, most of these methods do not explore semantic relationships of groupings, which could help better illuminate the groupings of PubMed abstracts.
To address this issue, we proposed an ontological clustering method called GOClonto for conceptualizing PubMed abstracts.
GOClonto uses Latent Semantic Analysis (LSA) and Gene Ontology (GO) to identify key gene-related concepts and their relationships as well as allocate PubMed abstracts based on these key gene-related concepts.