Particle Swarm Optimization as a New Measure of Machine Translation Efficiency

2018

The present work proposes a new approach to measuring efficiency of evolutionary algorithm-based Machine Translation. We implement some attributes of evolutionary algorithms performing cosine similarity objective function of a Particle Swarm Optimization (PSO) algorithm then, we evaluate an English text set for translation precision into the Spanish text as a simulated benchmark, and explore the backward process. Our results show that PSO algorithm can be used for translation of multiple language sentences with one identifier only, in other words the technology presented is language-pair independent. Specifically, we indicate that our cosine similarity objective function improves the veloci…