websvm
Websvm is web based implementation of LIBSVM. The site helps to train and test classification cases using SVM. The scripts access actual LIBSMV commands so full potential of the main library is inherent.
The trained model is an standard model file and can be exported to use it in Matlab and other applications that officially support LIBSVM library.
Websvm for now requires users to prepare their feature file in SVM format or sparse matrix format for training or testing purposes. Once users have a feature dataset, training SVM model, testing and generating classification are just few button clicks away.