Fig. 1From: Tackling the challenges of matching biomedical ontologiesFlowchart representation of AML’s pipeline. The pipeline is divided into three stages: ontology loading, where input or background knowledge (BK) ontologies are parsed and loaded into AML’s data structures; ontology matching, where matching algorithms generate mapping candidates which are combined into a preliminary alignment; and filtering, where problem-causing mappings are removed from the preliminary alignment to produce a final alignmentBack to article page