In reverse engineering, parsing may be partially done to extract lightweight source models. Parsing code containing preprocessing directives, syntactical errors and embedded langu...
We present a new grammar formalism for parsing with freer word-order languages, motivated by recent linguistic research in German and the Slavic languages. Unlike CFGs, these gram...
—Discovering program behaviors and functionalities can ease program comprehension and verification. Existing program analysis approaches have used text mining algorithms to infer...
Despite its state-of-the-art performance, the Data Oriented Parsing (DOP) model has been shown to suffer from biased parameter estimation, and the good performance seems more the ...
This paper shows how to formally characterize language learning in a finite parameter space as a Markov structure, hnportant new language learning results follow directly: explici...