Strategy for CFG Induction using Genetic Algorithm

Inductive Inference is the process of making generalizations from sample. The present problem is an inquiry regarding finding new strategy with use of an evolutionary approach, more specifically Genetic Algorithm, to perform context-free grammar induction. The work includes an effort towards determining the optimal structuring element to be used along with GA for the induction of Context Free Grammar. Some findings of the work are related to investigate the possibility of handling premature convergence problem in genetic algorithm with the application to grammar induction process. The work demonstrated the grammar induction process using proposed sequential structuring element with suggestion of the new mutation operator and the evaluation of the resulted grammar done by Push Down Automate simulator. The proposed technique is experimented and compared with the several existing techniques in the literature.