Please enter one sentence into the text box below and push the button. Even if you enter more than one sentence, only the first sentence will be parsed. Please let us know if you need to parse whole texts.

All modules are trained on the Penn Treebank II Wall Street Journal Corpus. Therefore, they work best with sentences of a similar kind. New sample sentences can be found at The New York Times on the Web.

This Shallow Parser for English consists of the following modules:

• The tokenizer splits punctuation marks from words.
• The tagger assigns one part-of-speech tag to each word.
• The chunker finds non-overlapping sequences of words (=chunks). Words inside a chunk closely belong together.
• The relation finder determines which chunk has which grammatical relation to which verbal chunk.

The latter three steps use the TiMBL Memory-Based Learning software package, which can be downloaded here.

Copyright ILK 2003

Publications about the Memory-Based Chunker and Grammatical Relation Finder:

• Memory-Based Grammatical Relation Finding. Sabine Buchholz. PhD thesis, Tilburg Universiteit, 2002.
• Memory-Based Shallow Parsing. Walter Daelemans, Sabine Buchholz, Jorn Veenstra. In: Proceedings of CoNLL-99, Bergen, Norway, June 12, 1999.
• Cascaded Grammatical Relation Assignment. Sabine Buchholz, Jorn Veenstra, Walter Daelemans. In: Proceedings of EMNLP/VLC-99, University of Maryland, USA, June 21-22, 1999.
• Memory-Based Text Chunking. Jorn Veenstra. In: Proceedings of ACAI, Chania, Greece, 1999.
• Fast NP chunking using memory-based learning techniques. Jorn Veenstra. In: F. Verdenius and W. van den Broek (Eds), Proceedings of Benelearn 1998, Wageningen, the Netherlands, pp. 71-79, 1998.
• Distinguishing complements from adjuncts using memory-based learning. Sabine Buchholz. In: B. Keller (Ed.), Proceedings of the ESSLLI-98 Workshop on Automated Acquisition of Syntax and Parsing, pp. 41-48, 1998.

Contact Walter Daelemans or Jorn Veenstra for more information about the chunker, and Walter Daelemans or Sabine Buchholz for more information about the grammatical relation finder.

Publications about the MBT tagger:

• Recent Advances in Memory-Based Part-of-Speech Tagging. Jakub Zavrel and Walter Daelemans. In: Actas del VI Simposio Internacional de Comunicacion Social, Santiago de Cuba, pp. 590-597, 1999.
• MBT: A Memory-Based Part of Speech Tagger-Generator. Walter Daelemans, Jakub Zavrel, Peter Berck and Steven Gillis. In: E. Ejerhed and I. Dagan (Eds) Proceedings of the Fourth Workshop on Very Large Corpora, Copenhagen, Denmark, pp. 14-27, 1996.

Contact Walter Daelemans or Jakub Zavrel for more information about the tagger.
Try also some taggers for other languages.

.