{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,3,12]],"date-time":"2025-03-12T04:37:32Z","timestamp":1741754252931,"version":"3.38.0"},"reference-count":0,"publisher":"University of Bor\u00e5s","issue":"iConf","license":[{"start":{"date-parts":[[2025,3,11]],"date-time":"2025-03-11T00:00:00Z","timestamp":1741651200000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IR"],"abstract":"<jats:p>Introduction. A vast amount of scholarly work is published daily, yet much of it remains inaccessible to the general public due to dense jargon and complex language. We introduce a reinforcement learning approach that fine-tunes a language model to rewrite scholarly abstracts into more comprehensible versions.\r\nMethod. Our approach utilises a carefully balanced combination of word- and sentence-level accessibility rewards to guide the language model in substituting technical terms with more accessible alternatives, a task which models supervised fine-tuned or guided by conventional readability measures struggle to accomplish.\r\nAnalysis. We evaluate our model\u2019s performance through readability metrics, factual accuracy assessments and language quality measurements, comparing results against supervised fine-tuning baselines.\r\nResults. Our best model adjusts the readability level of scholarly abstracts by approximately six US grade levels\u2014in other words, from a postgraduate to a high school level. This translates to roughly a 90% relative improvement over the supervised fine-tuning baseline, while maintaining factual accuracy and high-quality language.\r\nConclusions. We envision our work as a step toward bridging the gap between scholarly research and the general public, particularly younger readers, and those without a college degree.<\/jats:p>","DOI":"10.47989\/ir30iconf47530","type":"journal-article","created":{"date-parts":[[2025,3,11]],"date-time":"2025-03-11T13:54:59Z","timestamp":1741701299000},"page":"203-218","source":"Crossref","is-referenced-by-count":0,"title":["Improving scholarship accessibility with reinforcement learning"],"prefix":"10.47989","volume":"30","author":[{"given":"Haining","family":"Wang","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jason","family":"Clark","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hannah","family":"McKelvey","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Leila","family":"Sterman","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Gao","family":"Zheng","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zuoyu","family":"Tian","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiaozhong","family":"Liu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"27919","published-online":{"date-parts":[[2025,3,11]]},"container-title":["Information Research an international electronic journal"],"original-title":[],"link":[{"URL":"https:\/\/publicera.kb.se\/ir\/article\/download\/47530\/36955","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/publicera.kb.se\/ir\/article\/download\/47530\/36955","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,11]],"date-time":"2025-03-11T13:56:28Z","timestamp":1741701388000},"score":1,"resource":{"primary":{"URL":"https:\/\/publicera.kb.se\/ir\/article\/view\/47530"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,3,11]]},"references-count":0,"journal-issue":{"issue":"iConf","published-online":{"date-parts":[[2025,3,11]]}},"URL":"https:\/\/doi.org\/10.47989\/ir30iconf47530","relation":{},"ISSN":["1368-1613"],"issn-type":[{"value":"1368-1613","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,3,11]]}}}