[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-1-en-105":3,"doc-seo-238321-105":53,"doc-detail-238321-en":126},{"code":4,"msg":5,"data":6},0,"success",[7,14,19,24,29,34,39,44,49],{"id":8,"doc_module":9,"doc_module_name":10,"category_name":11,"show_sort_weight":12,"slug":13},11,1,"Template","Presentations",90,"presentations",{"id":15,"doc_module":9,"doc_module_name":10,"category_name":16,"show_sort_weight":17,"slug":18},12,"Resumes",80,"resumes",{"id":20,"doc_module":9,"doc_module_name":10,"category_name":21,"show_sort_weight":22,"slug":23},14,"Invoices",70,"invoices",{"id":25,"doc_module":9,"doc_module_name":10,"category_name":26,"show_sort_weight":27,"slug":28},15,"Posters",60,"posters",{"id":30,"doc_module":9,"doc_module_name":10,"category_name":31,"show_sort_weight":32,"slug":33},16,"Social Media",50,"social-media",{"id":35,"doc_module":9,"doc_module_name":10,"category_name":36,"show_sort_weight":37,"slug":38},17,"Forms",40,"forms",{"id":40,"doc_module":9,"doc_module_name":10,"category_name":41,"show_sort_weight":42,"slug":43},18,"Letters",30,"letters",{"id":45,"doc_module":9,"doc_module_name":10,"category_name":46,"show_sort_weight":47,"slug":48},21,"Paper Templates",5,"papers-templates",{"id":50,"doc_module":9,"doc_module_name":10,"category_name":51,"show_sort_weight":4,"slug":52},158,"General","general-158",{"code":4,"msg":54,"data":55},"ok",{"site_id":56,"language":57,"slug":58,"title":59,"keywords":60,"description":61,"schema_data":62,"social_meta":119,"head_meta":121,"extra_data":123,"updated_unix":125},105,"en","from-polarity-to-intensity-mining-morality-from-semantic-space","From Polarity to Intensity - Mining Morality from Semantic Space","","Most computational morality research focuses on moral polarity, distinguishing right from wrong, but polarity labels omit degree and intensity information. This paper introduces MORALSCORE, a weakly supervised framework that measures moral intensity from text using only polarity labels. It captures latent moral signals from both word-level and sentence-level semantics in a semantic space. Experiments with defined evaluation metrics show strong results on automatic scoring and human evaluations.",{"@graph":63,"@context":118},[64,80,101],{"@type":65,"itemListElement":66},"BreadcrumbList",[67,71,74,77],{"item":68,"name":69,"@type":70,"position":9},"https://docshare.wps.com","Home","ListItem",{"item":72,"name":10,"@type":70,"position":73},"https://docshare.wps.com/template/",2,{"item":75,"name":51,"@type":70,"position":76},"https://docshare.wps.com/template/general/",3,{"item":78,"name":59,"@type":70,"position":79},"https://docshare.wps.com/template/from-polarity-to-intensity-mining-morality-from-semantic-space/238321/",4,{"url":78,"name":59,"@type":81,"image":82,"author":87,"headline":59,"publisher":90,"fileFormat":93,"inLanguage":57,"description":61,"dateModified":94,"datePublished":95,"encodingFormat":93,"isAccessibleForFree":96,"interactionStatistic":97},"DigitalDocument",{"url":83,"@type":84,"width":85,"height":86},"https://docshare.wps.com/thumbnails/from-polarity-to-intensity-mining-morality-from-semantic-space/238321.png","ImageObject",442,249,{"name":88,"@type":89},"Ophelia","Person",{"url":68,"name":91,"@type":92},"DocShare","Organization","application/pdf","2026-09-21","2026-09-11",true,{"@type":98,"interactionType":99,"userInteractionCount":73},"InteractionCounter",{"@type":100},"ViewAction",{"@type":102,"mainEntity":103},"FAQPage",[104,110,114],{"name":105,"@type":106,"acceptedAnswer":107},"What limitation do existing computational morality methods have?","Question",{"text":108,"@type":109},"They mainly predict discrete moral polarity and do not provide intensity or degree information. This makes the modeling less informative for how severe an action is perceived.","Answer",{"name":111,"@type":106,"acceptedAnswer":112},"How does MORALSCORE compute moral intensity?",{"text":113,"@type":109},"MORALSCORE is weakly supervised and outputs a numerical moral intensity score for action-consequence pairs. It detects latent moral information from word to sentence semantics, then combines action and consequence via an explicit score combiner.",{"name":115,"@type":106,"acceptedAnswer":116},"Why are polarity labels sufficient for MORALSCORE training?",{"text":117,"@type":109},"The framework only requires moral polarity labels, which are more robust and easier to obtain than precise intensity annotations. This reduces reliance on subjective numerical intensity labeling.","https://schema.org",{"og:url":78,"og:type":120,"og:title":59,"og:site_name":91,"og:description":61},"article",{"robots":122,"canonical":78},"index,follow",{"doc_id":124,"site_id":56},238321,1789976625,{"code":4,"msg":5,"data":127},{"doc_id":124,"user_id":128,"nickname":88,"user_avatar":129,"doc_module":9,"category_id":50,"category_name":51,"doc_title":59,"doc_description":61,"doc_content":130,"file_id":131,"file_url":132,"file_type":133,"file_size":134,"view_count":73,"is_deleted":4,"is_public":9,"is_downloadable":9,"audit_status":9,"page_count":135,"language":136,"language_code":57,"site_id":56,"html_lang":57,"table_of_contents":137,"faqs":138,"seo_title":139,"seo_description":61,"update_tm":140,"read_time":47},7971461741311,"https://ap-avatar.wpscdn.com/avatar/74000253aff267980c6?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779345379180704826","From Polarity to Intensity: Mining Morality from Semantic Space  \nChunxu Zhao, Pengyuan Liu, Dong YuB  \nSchool of Information Science, Beijing Language and Culture University, China  \n[chunxu1212@gmail.com](chunxu1212@gmail.com), [liupengyuan@pku.edu.cn](liupengyuan@pku.edu.cn), [yudong_blcu@126.com](yudong_blcu@126.com)  \nAbstract  \nMost works on computational morality focus on moral polarity recognition, i.e., distinguishing right from wrong. However, a discrete polarity label is not informative enough to reflect morality as it does not contain any degree or intensity information. Existing approaches to compute moral intensity are limited to word-level measurement and heavily rely on human labelling. In this paper, we propose MORALSCORE, a weakly-supervised framework1 that can automatically measure moral intensity from text. It only needs moral polarity labels, which are more robust and easier to acquire. Besides, the framework can capture latent moral information not only from words but also from sentence-level semantics which can provide a more comprehensive measurement. To evaluate the performance of our method, we introduce a set of evaluation metrics and conduct extensive experiments. Results show that our method achieves good performance on both automatic and human evaluations.  \n1 Introduction  \nMoral intensity is a degree of feeling that a person has about a behaviour (Barnett, 2001) . As shown in Figure 1, although speeding on streets and killing a child are both immoral, the latter is more severe in most people’s perception. Understanding the above difference is an ability that humans have gradually developed in everyday life. It affects individuals’ ethical judgments and reflects the ideology of our society (Jones, 1991) . As AI gets ever more involved in people’s lives, it has become increasingly important for machine to acquire this ability and behave ethically. Researchers have studied the problem from early rule-based methods to today’s deep learning-based paradigms (Yu et al., 2018 ; Hendrycks et al., 2021) . It remains a fundamental but unsolved problem in computational morality (Moor, 2006) .  \n1 [https://github.com/blcunlp/MoralScore](https://github.com/blcunlp/MoralScore)  \nText Polarity Intensity  \nTom kills a child.  \n\n| 0 | 0.1 |\n| --- | --- |\n\nJoe is speeding on city streets.  \n\n| 0 | 0.4 |\n| --- | --- |\n\nIvy puts the rubbish in the dustbin.  \n\n| 1 | 0.7 |\n| --- | --- |\n\nBob donates 1M$ to charity.  \n\n| 1 | 0.9 |\n| --- | --- |\n\nFigure 1: Moral Intensity Example. From the numerical measurement of morality, both moral polarity and its degree can be reflected.  \nPrevious work in the NLP community often treats this problem as a supervised text classification task, i.e., judging the moral polarity for a text (Xie et al., 2020 ; Nahian et al., 2021) . This way of modelling morality is inadequate because it oversimplifies morality into a Bernoulli distribution, i.e., being only moral or immoral. We model morality into a continuous distribution by introducing moral intensity to include degree information. Computing moral intensity is challenging in two aspects: 1) In supervised settings, unlike labelling moral polarity, building a large corpus with precise intensity values is time consuming and prone to subjectivity. 2) In unsupervised settings, there is no direct link between text and moral intensity. Even when moral polarity labels are available, building such link is nontrivial because the binary labels do not reflect any information about moral intensity. To address these challenges, we propose MORALSCORE, a weakly-supervised framework that outputs a numerical value as the measurement of moral intensity for action-consequence pairs. The framework contains two parts. The first part is a semantic-aware moral detector, which measures moral intensity by detecting latent moral information from word to sentence level in semantic space. This incremental computing process can provide a comprehensive measurement o","cbCaioHwJFtDPoJI","https://ap.wps.com/l/cbCaioHwJFtDPoJI","pdf",1001279,13,"English","# Abstract\n# Introduction\n## Moral intensity definition and challenges\n## MORALSCORE framework overview\n# Semantic-Aware Moral Detector\n## Word-Level Self Scoring\n## Sentence-Level Interactive Scoring","[{\"question\":\"What limitation do existing computational morality methods have?\",\"answer\":\"They mainly predict discrete moral polarity and do not provide intensity or degree information. This makes the modeling less informative for how severe an action is perceived.\"},{\"question\":\"How does MORALSCORE compute moral intensity?\",\"answer\":\"MORALSCORE is weakly supervised and outputs a numerical moral intensity score for action-consequence pairs. It detects latent moral information from word to sentence semantics, then combines action and consequence via an explicit score combiner.\"},{\"question\":\"Why are polarity labels sufficient for MORALSCORE training?\",\"answer\":\"The framework only requires moral polarity labels, which are more robust and easier to obtain than precise intensity annotations. This reduces reliance on subjective numerical intensity labeling.\"}]","From Polarity to Intensity - Mining Morality from Semantic Space | PDF",1789133800]