A Study on the Best Way to Compress Natural Language Processing Models | IEEE Conference Publication | IEEE Xplore