机器学习预测目击者声音的准确性

IF 1.2 3区 心理学 Q4 PSYCHOLOGY, SOCIAL
Philip U. Gustafsson, Tim Lachmann, Petri Laukka
{"title":"机器学习预测目击者声音的准确性","authors":"Philip U. Gustafsson, Tim Lachmann, Petri Laukka","doi":"10.1007/s10919-024-00474-9","DOIUrl":null,"url":null,"abstract":"<p>An important task in criminal justice is to evaluate the accuracy of eyewitness testimony. In this study, we examined if machine learning could be used to detect accuracy. Specifically, we examined if support vector machines (SVMs) could accurately classify testimony statements as correct or incorrect based purely on the nonverbal aspects of the voice. We analyzed 3,337 statements (76.61% accurate) from 51 eyewitness testimonies along 94 acoustic variables. We also examined the relative importance of each of the acoustic variables, using Lasso regression. Results showed that the machine learning algorithms were able to predict accuracy between 20 and 40% above chance level (AUC = 0.50). The most important predictors included acoustic variables related to the amplitude (loudness) of speech and the duration of pauses, with higher amplitude predicting correct recall and longer pauses predicting incorrect recall. Taken together, we find that machine learning methods are capable of predicting whether eyewitness testimonies are correct or incorrect with above-chance accuracy and comparable to human performance, but without detrimental human biases. This offers a proof-of-concept for machine learning in evaluations of eyewitness accuracy, and opens up new avenues of research that we hope might improve social justice.</p>","PeriodicalId":47747,"journal":{"name":"Journal of Nonverbal Behavior","volume":null,"pages":null},"PeriodicalIF":1.2000,"publicationDate":"2024-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Machine Learning Predicts Accuracy in Eyewitnesses’ Voices\",\"authors\":\"Philip U. Gustafsson, Tim Lachmann, Petri Laukka\",\"doi\":\"10.1007/s10919-024-00474-9\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>An important task in criminal justice is to evaluate the accuracy of eyewitness testimony. In this study, we examined if machine learning could be used to detect accuracy. Specifically, we examined if support vector machines (SVMs) could accurately classify testimony statements as correct or incorrect based purely on the nonverbal aspects of the voice. We analyzed 3,337 statements (76.61% accurate) from 51 eyewitness testimonies along 94 acoustic variables. We also examined the relative importance of each of the acoustic variables, using Lasso regression. Results showed that the machine learning algorithms were able to predict accuracy between 20 and 40% above chance level (AUC = 0.50). The most important predictors included acoustic variables related to the amplitude (loudness) of speech and the duration of pauses, with higher amplitude predicting correct recall and longer pauses predicting incorrect recall. Taken together, we find that machine learning methods are capable of predicting whether eyewitness testimonies are correct or incorrect with above-chance accuracy and comparable to human performance, but without detrimental human biases. This offers a proof-of-concept for machine learning in evaluations of eyewitness accuracy, and opens up new avenues of research that we hope might improve social justice.</p>\",\"PeriodicalId\":47747,\"journal\":{\"name\":\"Journal of Nonverbal Behavior\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":1.2000,\"publicationDate\":\"2024-09-09\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Nonverbal Behavior\",\"FirstCategoryId\":\"102\",\"ListUrlMain\":\"https://doi.org/10.1007/s10919-024-00474-9\",\"RegionNum\":3,\"RegionCategory\":\"心理学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"PSYCHOLOGY, SOCIAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Nonverbal Behavior","FirstCategoryId":"102","ListUrlMain":"https://doi.org/10.1007/s10919-024-00474-9","RegionNum":3,"RegionCategory":"心理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"PSYCHOLOGY, SOCIAL","Score":null,"Total":0}
引用次数: 0

摘要

刑事司法中的一项重要任务是评估目击证人证词的准确性。在本研究中,我们考察了机器学习是否可用于检测准确性。具体来说,我们研究了支持向量机(SVM)是否可以纯粹根据声音的非语言方面准确地将证词陈述分为正确或不正确。我们根据 94 个声音变量分析了 51 份目击证人证词中的 3,337 项陈述(准确率为 76.61%)。我们还使用 Lasso 回归分析了每个声音变量的相对重要性。结果表明,机器学习算法能够预测高于偶然水平 20% 到 40% 的准确率(AUC = 0.50)。最重要的预测因素包括与语音振幅(响度)和停顿时间有关的声学变量,振幅越大,预测的正确率越高,停顿时间越长,预测的错误率越高。综上所述,我们发现机器学习方法能够预测目击证人证词的正确与否,准确率高于偶然性,与人类的表现不相上下,但不会产生有害的人为偏差。这为机器学习评估目击证人的准确性提供了概念证明,并开辟了新的研究途径,我们希望这能改善社会公正。
本文章由计算机程序翻译,如有差异,请以英文原文为准。

Machine Learning Predicts Accuracy in Eyewitnesses’ Voices

Machine Learning Predicts Accuracy in Eyewitnesses’ Voices

An important task in criminal justice is to evaluate the accuracy of eyewitness testimony. In this study, we examined if machine learning could be used to detect accuracy. Specifically, we examined if support vector machines (SVMs) could accurately classify testimony statements as correct or incorrect based purely on the nonverbal aspects of the voice. We analyzed 3,337 statements (76.61% accurate) from 51 eyewitness testimonies along 94 acoustic variables. We also examined the relative importance of each of the acoustic variables, using Lasso regression. Results showed that the machine learning algorithms were able to predict accuracy between 20 and 40% above chance level (AUC = 0.50). The most important predictors included acoustic variables related to the amplitude (loudness) of speech and the duration of pauses, with higher amplitude predicting correct recall and longer pauses predicting incorrect recall. Taken together, we find that machine learning methods are capable of predicting whether eyewitness testimonies are correct or incorrect with above-chance accuracy and comparable to human performance, but without detrimental human biases. This offers a proof-of-concept for machine learning in evaluations of eyewitness accuracy, and opens up new avenues of research that we hope might improve social justice.

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
Journal of Nonverbal Behavior
Journal of Nonverbal Behavior PSYCHOLOGY, SOCIAL-
CiteScore
4.80
自引率
9.50%
发文量
27
期刊介绍: Journal of Nonverbal Behavior presents peer-reviewed original theoretical and empirical research on all major areas of nonverbal behavior. Specific topics include paralanguage, proxemics, facial expressions, eye contact, face-to-face interaction, and nonverbal emotional expression, as well as other subjects which contribute to the scientific understanding of nonverbal processes and behavior.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信