java - Does MaltParser actually provide an option for returning probabilities of Parse trees? -

while looking @ source code of malt parser has class liblinear.java(jar file) , calls java version of liblinear toolkit; don't find option/way return probability despite information that, in principle training model using liblinear(by default in malt parser) logistic regression(-s 0) should produce probability score of parsed trees.

the main concern is: integration of liblinear , malt parser working smoothly without affecting each other expected operations?

working separately liblinear give me probability output datasets.

liblinear-train -s 0 train_scale  

//training data using logistic regression model

liblinear-predict -b 1 test_scale train_scale.model test_scale_output  

//labels , classes , probability outputs. here -b 1 extract out probabilities of each datasets.

reference: https://stackoverflow.com/questions/28791352/how-to-get-probability-score-of-parsed-sentences-using-malt-parser

malt parser works based on transition system , 2 or 3 stacks. @ each step transition predicted using liblinear or libsvm. input these models composed of in stacks , current state of machine. making decision @ 1 step affects rest of possible decisions. compute probability of tree require compute aggregated probabilites of trees (so sum 1), infeasable. compute trust score of tree, guess, or of particular arc, trust score, not probability. , afaik maltparser doesn't offer out of box. have alter source code, do-able think