{"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/quase-accurate-text-style-transfer-under","title":"QuaSE: Accurate Text Style Transfer under Quantifiable Guidance","arxiv_id":"1804.07007","date":"2018-04-19","proceeding":"EMNLP 2018 10","authors":["Yi Liao","Lidong Bing","Piji Li","Shuming Shi","Wai Lam","Tong Zhang"],"abstract":"We propose the task of Quantifiable Sequence Editing (QuaSE): editing an\ninput sequence to generate an output sequence that satisfies a given numerical\noutcome value measuring a certain property of the sequence, with the\nrequirement of keeping the main content of the input sequence. For example, an\ninput sequence could be a word sequence, such as review sentence and\nadvertisement text. For a review sentence, the outcome could be the review\nrating; for an advertisement, the outcome could be the click-through rate. The\nmajor challenge in performing QuaSE is how to perceive the outcome-related\nwordings, and only edit them to change the outcome. In this paper, the proposed\nframework contains two latent factors, namely, outcome factor and content\nfactor, disentangled from the input sentence to allow convenient editing to\nchange the outcome and keep the content. Our framework explores the\npseudo-parallel sentences by modeling their content similarity and outcome\ndifferences to enable a better disentanglement of the latent factors, which\nallows generating an output to better satisfy the desired outcome and keep the\ncontent. The dual reconstruction structure further enhances the capability of\ngenerating expected output by exploiting the couplings of latent factors of\npseudo-parallel sentences. For evaluation, we prepared a dataset of Yelp review\nsentences with the ratings as outcome. Extensive experimental results are\nreported and discussed to elaborate the peculiarities of our framework.","url_abs":"http://arxiv.org/abs/1804.07007v3","url_pdf":"http://arxiv.org/pdf/1804.07007v3.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":"quase-accurate-text-style-transfer-under","repo_url":"https://bitbucket.org/leoeaton/quase","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"disentanglement","task_name":"Disentanglement"},{"task_slug":"sentence","task_name":"Sentence"},{"task_slug":"style-transfer","task_name":"Style Transfer"},{"task_slug":"text-style-transfoer","task_name":"Text Style Transfer"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}