“…While entity recognition has been exploited as a powerful approach towards automatic NER retrieval, there has recently been an increased interest to find more complex structural information and more abundant knowledge in documents [ 9 ]. Hence, as a more recent development, moving beyond the purpose of entity recognition, the GENIA task in the BioNLP 2009/2011 shared tasks [ 10 , 11 ] was set to identify and extract nine biomedical events from GENIA-based corpus texts [ 12 ], including gene expression, transcription, protein catabolism, localization, binding, phosphorylation, regulation, positive regulation, and negative regulation.…”