Gold standard for English-Swedish Europarl data (GES)

Lars Ahrenberg - Linköping University, Department of Computer and Information Science

Maria Holmqvist - Linköping University, Department of Computer and Information Science

Reference corpus for word linking, divided into training data and test data. The sentences come from the English and Swedish parts of Europarl.

Data are created from the English-Swedish part of the Europarl corpus. For each sentence pair in the selected subset, token correspondences are stated as pairs of integral token identifiers