机器学习在社会科学中真的不安全、不负责任吗？累犯预测任务中的悖论与再思考

IF 1.8 4区社会学 Q2 CRIMINOLOGY & PENOLOGY

Asian Journal of Criminology Pub Date : 2024-06-26 DOI:10.1007/s11417-024-09429-x

Jianhong Liu, Dianshi Moses Li

{"title":"机器学习在社会科学中真的不安全、不负责任吗？累犯预测任务中的悖论与再思考","authors":"Jianhong Liu, Dianshi Moses Li","doi":"10.1007/s11417-024-09429-x","DOIUrl":null,"url":null,"abstract":"<div><p>The paper addresses some fundamental and hotly debated issues for high-stakes event predictions underpinning the computational approach to social sciences, especially in criminology and criminal justice. We question several prevalent views against machine learning and outline a new paradigm that highlights the promises and promotes the infusion of computational methods and conventional social science approaches.</p></div>","PeriodicalId":45526,"journal":{"name":"Asian Journal of Criminology","volume":"19 2","pages":"143 - 159"},"PeriodicalIF":1.8000,"publicationDate":"2024-06-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Is Machine Learning Really Unsafe and Irresponsible in Social Sciences? Paradoxes and Reconsideration from Recidivism Prediction Tasks\",\"authors\":\"Jianhong Liu, Dianshi Moses Li\",\"doi\":\"10.1007/s11417-024-09429-x\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><p>The paper addresses some fundamental and hotly debated issues for high-stakes event predictions underpinning the computational approach to social sciences, especially in criminology and criminal justice. We question several prevalent views against machine learning and outline a new paradigm that highlights the promises and promotes the infusion of computational methods and conventional social science approaches.</p></div>\",\"PeriodicalId\":45526,\"journal\":{\"name\":\"Asian Journal of Criminology\",\"volume\":\"19 2\",\"pages\":\"143 - 159\"},\"PeriodicalIF\":1.8000,\"publicationDate\":\"2024-06-26\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Asian Journal of Criminology\",\"FirstCategoryId\":\"90\",\"ListUrlMain\":\"https://link.springer.com/article/10.1007/s11417-024-09429-x\",\"RegionNum\":4,\"RegionCategory\":\"社会学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"CRIMINOLOGY & PENOLOGY\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Asian Journal of Criminology","FirstCategoryId":"90","ListUrlMain":"https://link.springer.com/article/10.1007/s11417-024-09429-x","RegionNum":4,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"CRIMINOLOGY & PENOLOGY","Score":null,"Total":0}

引用次数: 0

摘要

本文探讨了社会科学（尤其是犯罪学和刑事司法领域）计算方法在高风险事件预测方面的一些基本问题和激烈争论。我们对反对机器学习的几种流行观点提出质疑，并概述了一种新的范式，这种范式强调了计算方法和传统社会科学方法的前景，并促进了计算方法和传统社会科学方法的融合。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Is Machine Learning Really Unsafe and Irresponsible in Social Sciences? Paradoxes and Reconsideration from Recidivism Prediction Tasks

The paper addresses some fundamental and hotly debated issues for high-stakes event predictions underpinning the computational approach to social sciences, especially in criminology and criminal justice. We question several prevalent views against machine learning and outline a new paradigm that highlights the promises and promotes the infusion of computational methods and conventional social science approaches.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Asian Journal of Criminology CRIMINOLOGY & PENOLOGY-

CiteScore

3.00

自引率

10.50%

发文量

期刊介绍： Electronic submission now possible! Please see the Instructions for Authors. For general information about this new journal please contact the publisher at [welmoed.spahr@springer.com] The Asian Journal of Criminology aims to advance the study of criminology and criminal justice in Asia, to promote evidence-based public policy in crime prevention, and to promote comparative studies about crime and criminal justice. The Journal provides a platform for criminologists, policymakers, and practitioners and welcomes manuscripts relating to crime, crime prevention, criminal law, medico-legal topics and the administration of criminal justice in Asian countries. The Journal especially encourages theoretical and methodological papers with an emphasis on evidence-based, empirical research addressing crime in Asian contexts. It seeks to publish research arising from a broad variety of methodological traditions, including quantitative, qualitative, historical, and comparative methods. The Journal fosters a multi-disciplinary focus and welcomes manuscripts from a variety of disciplines, including criminology, criminal justice, law, sociology, psychology, forensic science, social work, urban studies, history, and geography.