From January to September 2026, the Institut für Translatologie gGmbH will run a cooperation project within the National Data Infrastructure consortium Text+, in collaboration with LinguisticBits. The goal of the project is to define a TEI-conform data format for multilingual projects. Two existing multilingual framenets, the lexicon of soccer jargon Kicktionary and the resource G-FOL for English-native learners of German, will be converted into this format.
Framenets are culture-sensitive concept databases. They describe knowledge structures (so-called frames) as well as linguistic forms that are connected to them. For instance, the G-FOL contains a frame entry Thinking: Opinion which illustrates that this knowledge structure represents a Thinking being that has an Opinion on a specific Topic which it holds to a certain Degree. The entry lists English and German linguistic forms such as verbs, nouns or sentence patterns by means of which we can express an opinion or the degree to which we hold it.
Framenets currently exist mainly in single languages. however, multilingual framenets and applications are becoming increasingly relevant, for instance, in order to create language learning resources, to assess the similarity of multilingual image descriptions or evaluate machine translation systems. The missing of a standard-conforming template for a data format has been a technical challenge for multilingual framenets to this date. The base format defined in the MuLingFrames project will alleviate this challenge. The project outcomes will be published under a non-commercial licence under completion.