LkbTop

Overview

The LKB system is a grammar and lexicon development environment for use with unification-based linguistic formalisms. While not restricted to HPSG, the LKB implements the DELPH-IN reference formalism of typed feature structures (jointly with other DELPH-IN software using the same formalism).

The primary documentation on the LKB is provided by the book [http://cslipublications.stanford.edu/lkb.html Implementing Typed Feature Structure Grammars]. Excerpts from the book provide an [http://cslipublications.stanford.edu/pdf/1575862603h.pdf tour of the LKB] (although see LkbInstallation for revised installation instructions) and the [http://cslipublications.stanford.edu/pdf/1575862603usersmanual.pdf user manual].

These pages are intended to provide documentation for aspects of the LKB not covered by the book, including recent developments. The use of a wiki forum is intended to enable LKB developers and users alike to contribute to the available on-line documentation. If you are new to the wiki paradigm, please familiarize yourself with some of the etiquettes and formatting guidelines before you start editing pages, create a wiki user name for yourself, and register on the server.

Background

The LKB has been in active use since around 1991, with a substantially new version in use from about 1997. These pages include documentation on the following topics:

• LkbCopyright: Copyright and Licensing;

• LkbInstallation: Obtaining, Installing and Running the LKB; ([https://depts.washington.edu/uwcl/twiki/bin/view.cgi/Main/KnoppixLKB Live CD: Knoppix LKB])

• LkbEmacs: Interfacing the LKB to the emacs(1) Editor;

• LkbLexDb: Using the lexical database machinery to build and maintain large lexicons;

• LkbTrollet: Trollet, a wrapper around LKB providing better multilingual support and a cleaner interface

• LkbUserGroup: Communication with LKB Developers and Other Users;

• LkbCourses: Information on the Use of the LKB in Teaching;

• LkbFaq: Frequently Asked Questions Related to Using the LKB;

• LkbUpdates: Changes in Recent Versions of the LKB; and

• LkbWishlist: suggestions/request for extensions to the LKB.

• LkbBugs: known issues with incorrect behavior in the LKB.

• LkbConventions: Conventions followed by many grammar writers.

• LkbGeneration: Information and notes on generating with the LKB

• [wiki:LkbLtdb Lexical Type Database]: A web interface to the lexicon (and treebanks). Thia is an add-on to the main LKB.

Other pages can be found by entering Lkb in the wiki search box. Please check this site regularly to see the LKB wiki grow. Or, if you are a knowledgeable LKB user, consider some volunteer work in creating up-to-date on-line documentation.

Grammars

Small sample grammars are downloadable as described in LkbInstallation. Larger scale resources which may be freely downloadable include:

• The [http://www.delph-in.net/matrix/ Matrix]: a starter kit for developing HPSG grammars.

• The [http://www.delph-in.net/erg/ English Resource Grammar] (ERG). Note that it requires at least 500 MBytes of RAM to run with the LKB. The latest version of the ERG may require the latest version of the LKB and not run on earlier versions.

• The [http://www.delph-in.net/jacy/ JACY] grammar: a broad-coverage HPSG implementation for Japanese.

• The [http://www.delph-in.net/mgrg/ Modern Greek Resource Grammar]: a medium-size grammar at this point (August 2005).

Projects

Projects which have made extensive use of the LKB include:

• [http://www.emmtee.net/ LOGON]

• [http://www.sciborg.org.uk SciBorg]

• [http://web.khu.ac.kr/~jongbok/projects/krg.html Korean Resource Grammar]

• [http://www.informatics.susx.ac.uk/research/nlp/cogent/ COGENT]

• [http://www.hcrc.ed.ac.uk/stanford/project-data2.cgi?datafile=data-28-03-03.tab&project=6 Robust Semantic Interpretation (ROSIE)]

• [http://www.project-deepthought.net/ DeepThought]

• [http://verbmobil.dfki.de/ VerbMobil]

• ACQUILEX I and II

Acknowledgements

Work on the LKB has been supported by a number of grants and funding sources, including:

• ACQUILEX I and II (funded by the European Commission);

• VerbMobil (funded by the German Ministry of Education and Science);

• The National Science Foundation under Grant No. IRI-9612682;

• The German Research Foundation (DFG) through the project PERFORM;

• Deep Thought funded under the 5th Framework Programme of the European Commission IST-2001-37836;

• NTT Communication Science Laboratories, Nippon Telegraph and Telephone Corporation (through joint research with CSLI);

• Engineering and Physical Sciences Research Council (EPSRC, UK). COGENT: Controlled generation of text and SciBorg: Extracting the Science from Scientific Publications;

• The Norwegian Research Council through the LOGON initiative;

• Boeing, grant to University of Cambridge.

Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the funding bodies.