Problems with Pruning in Automatic Creation of Semantic Valence Dictionary for Polish
TSD '09 Proceedings of the 12th International Conference on Text, Speech and Dialogue
The EM-based wordnet synsets annotation of NP/PP heads
LTC'09 Proceedings of the 4th conference on Human language technology: challenges for computer science and linguistics
Grouping alternating schemata in semantic valence dictionary of polish verbs
TSD'11 Proceedings of the 14th international conference on Text, speech and dialogue
SIIS'11 Proceedings of the 2011 international conference on Security and Intelligent Information Systems
Hi-index | 0.00 |
The ultimate goal of our work is to extend a syntactic valence dictionary of Polish verbs by adding some semantic information to verb arguments. This information consists of wordnet semantic categories of words. In order to provide syntactic slots of dictionary entries with lists of appropriate semantic categories of corresponding nouns, we need a treebank with all nouns semantically annotated with such categories, as both syntactic (i.e., argument structure) and semantic information is required.We aim here at Word Sense Disambiguation (WSD). To solve this task for our specific application, we adapt EM selection algorithm elaborated for extraction of syntactic valence frames.In the paper, the whole process of data processing is shown. The main focus is put on WSD task. Three versions of the EM selection algorithm are presented: the original one and its two modifications. Finally, the evaluation and comparison of the algorithms is performed.