This repository contains the evaluation sets used in <pre><code>A Benton, T Shi, O İrsoy, and I Malioutov."Weakly Supervised Headline Dependency Parsing". Findings of EMNLP. 2022. </code></pre> This dataset contains parse annotations for English news headlines and a script to produce conllu files joined with original headline text. Parse annotations are joined to the corresponding text by running: <pre> LDC_NYT_DIR="/PATH/TO/UNTARRED/LDC2008T19/" # path to untarred LDC2008T19 python build_eht.py --nyt_dir ${LDC_NYT_DIR} --num_proc 4</pre> This will download the Google sentence compression (GSC) dataset, and build conllu files for GSC examples. If you have the New York Times Annotated Corpus (LDC2008T19) untarred locally, this will also join annotations to the NYT examples (location passed via <code>--nyt_dir</code>). Increase the argument to <code>--num_procs</code> to process more shards from the NYT corpus in parallel and reduce build time. The above was tested with python 3.9.7. The EHT evaluation sets, with gold-annotated POS tags and dependency relations, are built as <code>EHT/gsc.test.conllu</code> and <code>EHT/nyt.test.conllu</code> Silver, projected, trees which we used to train and validate out models are built under <code>GSC_projected</code>. These are not gold parse trees (projected predictions from the article lead sentence), and are shared purely for reproducibility sake.