Integer Linear Programming has recently been used for decoding in a number of probabilistic models in order to enforce global constraints. However, in certain applications, such as non-projective dependency parsing and machine translation, the complete formulation of the decoding problem as an integer linear program renders solving intractable. We present an approach which solves the problem incrementally, thus we avoid creating intractable integer linear programs. This approach is applied to Dutch dependency parsing and we show how the addition of linguistically motivated constraints can yield a significant improvement over stateof-the-art. 1

In this project report we describe work in statistical parsing using the maximum entropy technique and the Alpino language analysis system for Dutch. A major difficulty in this domain is the lack of sucient corpus data available for training. Among other problems, this sparseness of data increases the danger of the model overfitting the training data, making it particularly important that the selection of statistical features upon which to base the model be optimal. To this end we have adapted the notion of feature merging, a means of constructing equivalence classes of statistical features based upon common elements within them. In spite of promising preliminary results, subsequent tests have not enabled us to conclude whether this approach helps the kind of models we are working with.

We investigate the factors that influence argument order variations in Dutch, focusing on ditransitive verbs. Evidence from grammaticality judgments is complemented with evidence from the distributions of the alternants in corpora of spoken and written Dutch. We find that while the NP/PP alternation is influenced by weight, the direct object shift (DOS) in the dative alternation is bound to certain types of pronouns. Additional evidence for our account for the DOS is found in the Dutch AcI construction. Our findings are modeled within the framework of Optimality Theoretic syntax, allowing for violable and ranked constraints, as well as a stochastic interpretation of the analysis. 1