The ACL OCL Corpus: advancing Open science in Computational Linguistics

We present a scholarly corpus from the ACL Anthology to assist Open scientific research in the Computational Linguistics domain, named as ACL OCL. Compared with previous ARC and AAN versions, ACL OCL includes structured full-texts with logical sections, references to figures, and links to a large knowledge resource (semantic scholar). ACL OCL contains 74k scientific papers, together with 210k figures extracted up to September 2022. To observe the development in the computational linguistics domain, we detect the topics of all OCL papers with a supervised neural model. We observe ''Syntax: Tagging, Chunking and Parsing'' topic is significantly shrinking and ''Natural Language Generation'' is resurging. Our dataset is open and available to download from HuggingFace in https://huggingface.co/datasets/ACL-OCL/ACL-OCL-Corpus.