Feature Selection Using NSGA-II for Event Extraction on Genetic and Molecular Mechanisms Involved in Plant Seed Development

Abstract

Molecular network structure to regulate plant seed development is very complex and to understand this from biomedical literature is a big challenge. Seed development is based on coordinated growth of different tissues, which are involved with complex genetics and environmental regulation. We develop a system for binary event extraction using statistical-, syntactic-, and dependency-based features. Experiments on the benchmark datasets of BioNLP-2016 SeeDev shared task show the recall, precision, and F-score values of 0.517, 0.399, and 0.451, respectively.