{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/a-mention-ranking-model-for-abstract-anaphora","title":"A Mention-Ranking Model for Abstract Anaphora Resolution","arxiv_id":"1706.02256","date":"2017-06-07","proceeding":"EMNLP 2017 9","authors":["Ana Marasović","Leo Born","Juri Opitz","Anette Frank"],"abstract":"Resolving abstract anaphora is an important, but difficult task for text\nunderstanding. Yet, with recent advances in representation learning this task\nbecomes a more tangible aim. A central property of abstract anaphora is that it\nestablishes a relation between the anaphor embedded in the anaphoric sentence\nand its (typically non-nominal) antecedent. We propose a mention-ranking model\nthat learns how abstract anaphors relate to their antecedents with an\nLSTM-Siamese Net. We overcome the lack of training data by generating\nartificial anaphoric sentence--antecedent pairs. Our model outperforms\nstate-of-the-art results on shell noun resolution. We also report first\nbenchmark results on an abstract anaphora subset of the ARRAU corpus. This\ncorpus presents a greater challenge due to a mixture of nominal and pronominal\nanaphors and a greater range of confounders. We found model variants that\noutperform the baselines for nominal anaphors, without training on individual\nanaphor data, but still lag behind for pronominal anaphors. Our model selects\nsyntactically plausible candidates and -- if disregarding syntax --\ndiscriminates candidates using deeper features.","url_abs":"http://arxiv.org/abs/1706.02256v2","url_pdf":"http://arxiv.org/pdf/1706.02256v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"a-mention-ranking-model-for-abstract-anaphora","repo_url":"https://github.com/amarasovic/neural-abstract-anaphora","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"abstract-anaphora-resolution","task_name":"Abstract Anaphora Resolution"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"sentence","task_name":"Sentence"},{"task_slug":"model","task_name":"model"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/abstract-anaphora-resolution-on-the-arrau","task":"Abstract Anaphora Resolution","dataset":"The ARRAU Corpus","model":"MR-LSTM","rank_in_archive_order":1,"of":1,"metrics":{"Average Precision":"43.83"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1706.02256","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}