Generating the Persian Constituency Treebank in an Automatic Converting Method

Treebanks is one of important and useful resource in Natural Language Processing tasks. Dependency and phrase structures are two famous kinds of treebanks. There have already made many efforts to convert dependency structure to phrase structure. In this paper we study an approach to convert dependency structure to phrase structure because of lack of a big phrase structure Treebank in Persian. Also we study the algorithm’s errors and propose a solution to solve the problem and improve the quality of conversion process. The experiment results show that we can improve the quality of conversion, about 25.85 percent, in Persian and about 4.39 percent in English. With the help of the conversion algorithm and the dependency Treebank, we produce the phrase structure treebank and train a parser using the resulted treebank. Our parser output is about 21 percent, better than the same parser introduced as baseline.