Seminar für Sprachwissenschaft


WebLicht: Web-based LRT Services

WebLicht is a Service Oriented Architecture (SOA) for the creation of annotated text corpora. It has been under development since October 2008 in the D-SPIN and CLARIN-D project, together with some of the project partners. In 2011, WebLicht contained about 130 tools and resources for the creation, analysis, and visualization of annotated text corpora.


CASS is a finite-state parser which uses a cascade of finite-state automata. It is optimized for speed and memory.

CASS was developed by Steven Abney while he was in Tübingen. The latest version of this parser can be downloaded from his website.


BibTex-HTML Converter

Our bibTeX-XML-HTML package transforms a (potentially thematically structured) BibTeX bibliography into a set of HTML files to facilitate their publication on the web.

The BibTeX-XML-HTML converter tools will first transform your BibTeX file into an XML file. The advantage of the XML format is that it is standardized, platform independent and universal. Furthermore, it is easily edited and managed with XML tools and translated into HTML (or text files) with the XSLT stylesheet language. Our tools will generate multiple HTML files out of the XML representation of the bibliography for publication on the web.

Contact: Frank Richter


FSA Utilities

The FSA Utilities are a collection of tools that construct finite automata from regular expressions, manipulate finite automata, visualise finite automata, and apply finite automata.

This toolbox was developed mainly by Gertjan van Noord at the University of Groningen in cooperation with Dale Gerdemann. 

Contact: Dale Gerdemann


LexParse is an easy-to-use, fast and flexible programme for parsing machine-readable dictionaries. It parses typesetting tapes or other text files by applying user-defined grammars. LexParse identifies single dictionary entries and recognizes their internal structure. LexParse is able to display the resulting parse trees in different user-definable formats (tree structure, SGML format, LaTex).



The TRALE system is an implementation platform for feature-structure based grammars. It is a combination of the experience from the development and application of our earlier systems for computing with HPSG-style grammars. The idea is to combine the advantage of efficiency of more traditional parsing systems with the goal of submitting a depiction, as true to form as possible, of theoretical HPSG grammars into computational implementation.

Contact: Frank Richter