Automatic noun phrases extraction using preliminary segmentation and CRF with semantic features

We consider the task of finding the borders of noun phrases (NP) that are actants of predicates. First, we make a preliminary segmentation of sentences to fragments that contain NPs. Second, we use CRF to find the borders of NPs inside the fragments. Data from the knowledge base and information about named entities found in the text are used as features for machine learning. We present the results of our experiment and discuss future work. (in Russian).