In 2008 ISO Technical Committee 37, Terminology and other language and content resources, has published ISO 24613 Lexical Markup Framework (LMF; see also The LMF standard provides an abstract lexicon model based on a core model and several optional extensions, e.g., morphology, NLP, or semantics.  The abstract model is expressed in UML (Unified Modeling Language) using classes and relationships between them. Users of LMF need to instantiate the abstract model by adorning the UML class model with feature structures referring to data categories taken from the ISOcat Data Category Registry. Only the abstract LMF lexicon UML model is normative. The LMF XML DTD provided by the standard is informative.

TLA and LMF have a long shared history. Work on the web-based LEXUS lexicon tool started in the EU LIRICS project in which also the first drafts of the LMF standard were written. After the end of the LIRICS project the development of LEXUS and LMF drifted apart. LEXUS became a free form lexicon tool, i.e., users are not limited to instantiating the abstract LMF model but can create almost any lexicon structure they want. However, basic LMF features, e.g., in particular the link with ISOcat, stayed present.

The aim of the just finished DFG/NEH RELISH project was to foster interchange between LMF, used primarily in Europe, and LIFT as used primarily in the United States. This resulted in a more modular XML schema for LMF named RELISH LMF and a specific set of RELISH LL LMF modules providing an interoperable core of both LMF and LIFT. Stylesheets to round trip lexicons based on this interoperable core between the LMF-based LEXUS TLA tool and the LIFT-based LEGO LinguistList project are becoming available.

Based on the results of RELISH first important steps are undertaken to align LEXUS and LMF again. LEXUS 3 introduced templates to assist users in creating a lexicon structure. LEXUS 3.1 will extend this further with import and export functions, e.g., from and to RELISH (LL) LMF, and thus improve LEXUS’ support for LMF.