Interlingua Development and Testing through Semantic Annotation of Multilingual Text Corpora

This paper describes a multi-site project to anno- tate the interlingual content of six sizable bilin- gual parallel corpora. The project addresses several principal problems in parallel: specifica- tion of interlingua content and notation, develop- ment of reliable annotation methods, and evaluation of annotated corpora. As a by-product, a growing corpus of annotated texts is being pro- duced, which may eventually be useful for ma- chine learning of semantics-based processing.