About TypeCraft

under construction :=)

Background

Since the mid eighties, a group of researchers and students at the Linguistics Department at the Norwegian University of Science and Technology has explored the use of formal and shallow linguistic methods for natural language applications. The formalization and encoding of morpho-syntactic and semantic information, both at lexical and phrasal level, has been a central theme for this group with in 2003 took the name LinLab.

At present LinLab has two focal areas: Grammar Engineering and Language Documentation.

Grammar Engineering at LingLab

In Grammar Engineering the main application developed by LinLab is the Norwegian computational grammar NorSource (Lars Hellan, Dorothee Beermann and Ben Waldron). Together with partners in the DELPH-IN network LinLab applies Head-Driven Phrase Structure Grammar (Pollard and Sag (1994) and Minimal Recursion Semantics (MRS)(Copestake et.al.2005) to advance deep natural language processing.

As part of this work Pavel Mihaylov developed for LingLab an LKB multi-script interface called Trollet.

A further effort to represent lexical and construction level information is the Construction Labeling Project Verbconstructions cross-linguistically - Introduction, a system for encoding construction types across languages (Lars Hellan).

TypeCraft

TypeCraft itself is a product of LingLab's effort in Language Documentation. A prototype of TypeCraft was developed in 2006 by Dorothee Beermann and Atle Prange at Businesscape. The present beta version of TypeCraft represents a joined effort by Pavel Mihaylov and Dorothee Beermann.

A short description of TypeCraft

TypeCraft is an INTERLINEAR GLOSSING EDITOR; the user adds linguistic annotation to written material which is stored in a relational database from where it can be retrieved using multiple views.

Texts as well as annotations are in Unicode. Annotated data can be exported to standard text editors (WORD, Open Office and LaTex) as well as to XML format. TC has been designed for projects on minority languages, which are often projects joining linguists and native speakers. To avoid long training periods, TypeCraft features an intuitive user interface. Often essential for fieldwork conducted as part of a collaboration, TypeCraft allows wide and distributive usage.

TC uses PostgreSQL as database format. The database is hosted at a server owned by the Norwegian University of Science and Technology in Trondheim. The application is written in Java.

TypeCraft can be freely use online with Mozilla Firefox.