- Full text PDF available (3)
- This year (0)
- Last 5 years (3)
- Last 10 years (3)
Journals and Conferences
We present an approach to obtain language models from a tree corpus using probabilistic regular tree grammars (prtg). Starting with a prtg only generating trees from the corpus, the prtg is generalized step by step by merging nonterminals. We focus on bottom-up deterministic prtg to simplify the calculations.
We prove that the class of linear context-free tree languages is not closed under inverse linear tree homomorphisms. The proof is by contradiction: we encode Dyck words into a context-free tree language and prove that its preimage under a certain linear tree homomorphism cannot be generated by any context-free tree grammar. A positive result can still be… (More)
Encoding unranked trees to binary trees, henceforth called binarization, is an important method to deal with unranked trees. For each of three binarizations we show that weighted (ranked) tree automata together with the binarization are equivalent to weighted unranked tree automata; even in the probabilistic case. This allows to easily adapt training… (More)