Composition Models for the Representation and Semantic Interpretation of Nominal Compounds

Contains five datasets: two for German composition (de-nn-only-comp, de-mixed-comp), one for English composition (en-comp), one containing semantically annotated German compounds (de-nncom-sem) and one containing German compounds annotated for lexicalization (de-ulex). Includes also word vectors in four dimensions for German and English words and compounds.

word embeddings - distributional representations - word vectors - phd thesis - semantic annotations - semantic composition datasets - German compounds - English compounds