1 SVM is a classifier, and its accuracy and fitting degree have been improved very maturely, which is difficult to be improved by mathematical methods, but it is widely used. SVMRank seems to be the core algorithm of Netflix movie recommendation system, so you can learn about it.
2 boosting is an integrated algorithm combined with other algorithms. You can regard SVM as a weak learner to improve the accuracy of learning within its framework.
The algorithm of SVM itself has really improved and can be published in the highest level magazines. Although the two aspects I mentioned above are very simple, if you have experimental data to prove that it is no problem to publish core journals in China, I am also entangled in the paper. .