I have some URLs and i have to extract some features from these URLs to filter out some desired URLs. I have to give input those filtered URLs to SVM so as to classify those URLs as relevant and irrelevant. What will be the exact input to SVM?
You would probably need to use an XML reader and Natural Language Programming techniques to create features from the data which would be useful in determining relevancy. You would also have to label the URLs as relevant or irrelevant.
Both the class labels and the features would be the input into your [SVM].