Corpus Query Lingua Franca (CQLF)

Corpus Query Lingua Franca (CQLF) Piotr Bański1, Elena Frick1, Andreas Witt1, 2 1Institute for the German Language (IDS Mannheim) 2 Heidelberg University Email: {banski|frick|witt}@ids-mannheim.de Abstract The present paper describes Corpus Query Lingua Franca (ISO CQLF), a specification designed at ISO Technical Committee 37 Subcommittee 4 “Language resource management” for the purpose of facilitating the comparison of properties of corpus query languages. We overview the motivation for this endeavour and present its aims and its general architecture. CQLF is intended as a multi-part specification; here, we concentrate on the basic metamodel that provides a frame that the other parts fit in. Keywords: text corpora, query languages, ISO standards 1. Introduction The present paper describes Corpus Query Lingua Franca (ISO CQLF), a specification designed at ISO Technical Committee 37 Subcommittee 4 “Language resource management” for the purpose of facilitating the comparison of properties of corpus query languages. We overview the motivation for this endeavour and present its aims and its general architecture. CQLF is intended as a multi-part specification; here, we concentrate on the basic metamodel that provides a frame that the other parts fit in. 1.1. Aims and Motivation A lot of effort has been devoted to developing various corpus query languages (cQLs) in order to adequately satisfy information needs of corpus linguists and lexicographers. In practice, this large amount of different cQLs (see e.g. Clematide 2015) poses a serious challenge for researchers seeking to maximally exploit the advantages offered by contemporary language resources; this is true especially for those who do not have a background in computational linguistics and programming. Apart from learnability problems concerning some cQLs, the real problem is that a number of linguistic resources are accessible through drastically different corpus management systems which provide their own cQLs. Thus, mastering one or two cQLs is not enough to benefit from the entire palette of existing corpus linguistic software packages. Apart from that, as the respective fields develop, information needs grow and it may well happen that after having invested significant time and effort into learning and adopting a new cQL, a researcher will come to realise that the given cQL provides no way to further elaborate on the existing queries, and that another cQL has to be used. There exists a need for interoperability across corpus query systems, realized among others by abstracting away from individual cQLs and seeing how far their queries can be compared. The issue of general interoperability of language resources and applications has been addressed at various levels of abstraction by such initiatives as e.g. CLARIN “Federated Content Search”, ISOCat (now in redesign), or WSEP (Ide, 2013). With the purpose of enabling interoperability of corpus query tools (or at least gauging the extent to which such interoperability is possible), we have initiated the development of a standard for the presentation of features of corpus QLs, called Corpus Query Lingua Franca (CQLF), within ISO TC37 SC4 WG6. Currently at the Committee Draft stage, ISO CQLF (ISO/CD 24623-1) is the first standardization activity in the domain …

Content truncated to 500 words