{"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/better-rewards-yield-better-summaries","title":"Better Rewards Yield Better Summaries: Learning to Summarise Without References","arxiv_id":"1909.01214","date":"2019-09-03","proceeding":"IJCNLP 2019 11","authors":["Florian Böhm","Yang Gao","Christian M. Meyer","Ori Shapira","Ido Dagan","Iryna Gurevych"],"abstract":"Reinforcement Learning (RL) based document summarisation systems yield state-of-the-art performance in terms of ROUGE scores, because they directly use ROUGE as the rewards during training. However, summaries with high ROUGE scores often receive low human judgement. To find a better reward function that can guide RL to generate human-appealing summaries, we learn a reward function from human ratings on 2,500 summaries. Our reward function only takes the document and system summary as input. Hence, once trained, it can be used to train RL-based summarisation systems without using any reference summaries. We show that our learned rewards have significantly higher correlation with human ratings than previous approaches. Human evaluation experiments show that, compared to the state-of-the-art supervised-learning systems and ROUGE-as-rewards RL summarisation systems, the RL systems using our learned rewards during training generate summarieswith higher human ratings. The learned reward function and our source code are available at https://github.com/yg211/summary-reward-no-reference.","url_abs":"https://arxiv.org/abs/1909.01214v1","url_pdf":"https://arxiv.org/pdf/1909.01214v1.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":"better-rewards-yield-better-summaries","repo_url":"https://github.com/yg211/summary-reward-no-reference","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"better-rewards-yield-better-summaries","repo_url":"https://github.com/UKPLab/emnlp2019-summary-reward","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1909.01214","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1909.01214"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/UKPLab/emnlp2019-summary-reward","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/yg211/summary-reward-no-reference","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran_draft_wrong":2,"ran_honours":2,"unverified":10},"by_repo_kind":{"official":{"samples":9,"ran":0,"repositories":1},"listed":{"samples":5,"ran":4,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":5,"samples":[{"code_sha256_prefix":"7b6ce349a42f2945","entry":"build_model","repo":"UKPLab/emnlp2019-summary-reward","repo_kind":"listed","path":"rewarder.py","file_url":"https://github.com/UKPLab/emnlp2019-summary-reward/blob/HEAD/rewarder.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"7b6ce349a42f2945"}},{"code_sha256_prefix":"ceee2c8bc03cc5b0","entry":"parse_split_data","repo":"UKPLab/emnlp2019-summary-reward","repo_kind":"listed","path":"step2_train_rewarder.py","file_url":"https://github.com/UKPLab/emnlp2019-summary-reward/blob/HEAD/step2_train_rewarder.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ceee2c8bc03cc5b0"}},{"code_sha256_prefix":"d193c2053b5ce0cb","entry":"raw_bert_encoder","repo":"UKPLab/emnlp2019-summary-reward","repo_kind":"listed","path":"rewarder.py","file_url":"https://github.com/UKPLab/emnlp2019-summary-reward/blob/HEAD/rewarder.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"d193c2053b5ce0cb"}},{"code_sha256_prefix":"2cbfa7df6a518aa2","entry":"raw_bert_encoder","repo":"UKPLab/emnlp2019-summary-reward","repo_kind":"listed","path":"step1_encode_doc_summ.py","file_url":"https://github.com/UKPLab/emnlp2019-summary-reward/blob/HEAD/step1_encode_doc_summ.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2cbfa7df6a518aa2"}},{"code_sha256_prefix":"b8a6998c443cf95c","entry":"bleu","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"scorer/auto_metrics/metrics.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/scorer/auto_metrics/metrics.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b8a6998c443cf95c"}},{"code_sha256_prefix":"7c4d342f8afe60b2","entry":"countAppearance","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"scorer/data_helper/generate_samples.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/scorer/data_helper/generate_samples.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"7c4d342f8afe60b2"}},{"code_sha256_prefix":"c6edfba6734d0282","entry":"deep_pair_train","repo":"UKPLab/emnlp2019-summary-reward","repo_kind":"listed","path":"step2_train_rewarder.py","file_url":"https://github.com/UKPLab/emnlp2019-summary-reward/blob/HEAD/step2_train_rewarder.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c6edfba6734d0282"}},{"code_sha256_prefix":"b1505c96d52fa4ce","entry":"getPossibleSwaps","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"scorer/data_helper/generate_samples.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/scorer/data_helper/generate_samples.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b1505c96d52fa4ce"}},{"code_sha256_prefix":"08a56c9270b8138e","entry":"inHistory","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"scorer/data_helper/generate_samples.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/scorer/data_helper/generate_samples.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"08a56c9270b8138e"}},{"code_sha256_prefix":"b76aabfbdd5cd7dd","entry":"remove_spaces_lines","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"helpers/data_helpers.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/helpers/data_helpers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b76aabfbdd5cd7dd"}},{"code_sha256_prefix":"59b18c5647389fbf","entry":"sent2tokens","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"helpers/data_helpers.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/helpers/data_helpers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"59b18c5647389fbf"}},{"code_sha256_prefix":"ba627e31aa948470","entry":"sts_bert_encoder","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"compare_reward.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/compare_reward.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"ba627e31aa948470"}},{"code_sha256_prefix":"6abe2a89945836f0","entry":"sts_bert_rewarder","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"compare_reward.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/compare_reward.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6abe2a89945836f0"}},{"code_sha256_prefix":"ff0f39e164572c49","entry":"text_normalization","repo":"yg211/summary-reward-no-reference","repo_kind":"official","path":"helpers/data_helpers.py","file_url":"https://github.com/yg211/summary-reward-no-reference/blob/HEAD/helpers/data_helpers.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"ff0f39e164572c49"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}