Skip navigation.
Home
Semantic Software Lab
Concordia University
Montréal, Canada

Multi-lingual Noun Phrase Extractor (MuNPEx)

Printer-friendly versionPrinter-friendly versionPDF versionPDF version

1. Overview

The Multi-Lingual Noun Phrase Extractor (MuNPEx) is a noun phrase (NP) chunker component developed for the GATE architecture, implemented in JAPE. It is fast, robust, customizable, well-tested and currently supports English, German, and French (with Spanish in beta). It provides detailed features for each NP annotation, with DET (determiner), MOD/MOD2 (pre/post-head modifiers), and HEAD noun slots, as well as (optional) text offset information.

MuNPEx requires a part-of-speech (POS) tagger to work and can additionally use detected named entities (NEs) to improve chunking performance. For English, MuNPEx works with the ANNIE (Hepple) tagger that comes as part of the ANNIE system with GATE. French, German, and Spanish are based on the TreeTagger. Please read the included documentation (or source code) for more details. Also note that you will need at least GATE version 5.x to run MuNPEx, as it makes use of the new Kleene operators for JAPE.

MuNPEx in GATE Developer: Screenshot of MuNPEx-1.0 running on a German textMuNPEx in GATE Developer: Screenshot of MuNPEx-1.0 running on a German text

2. Installation

If you have GATE 7 or better, you can install MuNPEx directly from within GATE using the CREOLE Plugin Manager by selecting our Semantic Software Lab repository. The install package includes documentation, the main and supplementary JAPE grammars, the example pipeline, and some additional information. You can also download the install package manually (but the recommended way of installation is to use the GATE plugin manager through the GATE Developer GUI).

3. Version history

  • 1.2: 24.04.2012. Minor bugfix to English chunker
  • 1.1: 12.02.2012. Repackaged for new GATE 7 Plugin Manager. Added example pipeline for English. Added wrappers to load MuNPEx grammars as PRs.
  • 1.0: 16.08.2010. More robust on malformed input. Optional grammars for adding a HEAD_LEMMA slot. DET/MOD/HEAD/MOD2 now stored as strings instead of Content objects. Code cleanup and tweaks.
  • 0.2: 03.03.2006. Preliminary Spanish support added. Renamed from "NPE" to "MuNPEx". Small cleanups. Number transducer added.
  • 0.1: 21.11.2005. Initial public release.

4. License

MuNPEx is distributed as free/open source software under the GNU GPL license v2.

5. Feedback

For questions, comments, etc., please use the Forum.

6. Acknowledgements

Thanks to Michelle Khalifé for helping out in developing the French version.