If this question is too general, please let me know, and I will revise it.
I am developing a program that measures all sorts of "features" of a sentence, everything from sentence length, to average word length, to lexical diversity. I'd also like to incorporate some features of the semantic graphs of the sentences.
What are some suggestions as to useful features of the graph to extract? I'm just looking for a handful, to get the ball rolling, and then develop them further for my specific needs.
Any reading suggestions would also be appreciated.