The Second Multilingual Surface Realisation Shared Task (SR'19) - Overview and Evaluation Results.

We report results from the SR’19 Shared\r\nTask, the second edition of a multilingual surface realisation task organised as part of the\r\nEMNLP’19 Workshop on Multilingual Surface Realisation. As in SR’18, the shared task\r\ncomprised two different tracks: (a) a Shallow\r\nTrack where the inputs were full UD structures\r\nwith word order information removed and tokens lemmatised; and (b) a Deep Track where\r\nadditionally, functional words and morphological information were removed. The Shallow\r\nTrack was offered in 11, and the Deep Track\r\nin three languages. Systems were evaluated\r\n(a) automatically, using a range of intrinsic\r\nmetrics, and (b) by human judges in terms of\r\nreadability and meaning similarity to a reference. This report presents the evaluation results, along with descriptions of the SR’19\r\ntracks, data and evaluation methods, as well as\r\nbrief summaries of the participating systems.\r\nFor full descriptions of the participating systems, please see the separate system reports\r\nelsewhere in this volume.