Comparison of Crossover Types to Build Improved Queries Using Adaptive Genetic Algorithm
Khaled Almakadmeh, Wafa’ Za’al Alma’aitah · 2017
This paper presents an information retrieval system that use genetic algorithm to improve information retrieval efficiency and vector space model to measure similarity between query and documents retrieved. Therefore, documents with high similarity to query retrieved first. Using the genetic algorithm, each query represented by a chromosome, these chromosomes are fed into genetic operator process: selection, crossover, and mutation until a query chromosome generated for document retrieval. The proposed approach is experimented using a data set of (242) proceedings abstracts collected from a Saudi Arabian national conference. Experimental results show that information retrieval with adaptive crossover probability set to two-point type crossover and roulette wheel as selection type yields the highest recall.