Extraction of text classification rules based on multi-population collaborative optimization

摘要

Most text classification methods are highly complicated on computation and can not be used on the occasion of classifying a large number of texts. A novel approach based on multi-population collaborative optimization was proposed for the extraction of text classification rules. The mutual information was applied to generate the initial populations and the multi-population collaborative optimization method was adopted to evolve the current population. Experimental results show that the number of classification rules is small, the accuracies of classification rules are high and the time of computation is short using this approach. And this approach is competent for processing the large-scale text datasets.