Neubauer Lilly, Straw Isabel, Mariconti Enrico, Tanczer Leonie Maria
University College London, Gower Street, London, WC1E 6BT UK.
J Fam Violence. 2023 Mar 21:1-20. doi: 10.1007/s10896-023-00517-7.
Computational text mining methods are proposed as a useful methodological innovation in Intimate Partner Violence (IPV) research. Text mining can offer researchers access to existing or new datasets, sourced from social media or from IPV-related organisations, that would be too large to analyse manually. This article aims to give an overview of current work applying text mining methodologies in the study of IPV, as a starting point for researchers wanting to use such methods in their own work.
This article reports the results of a systematic review of academic research using computational text mining to research IPV. A review protocol was developed according to PRISMA guidelines, and a literature search of 8 databases was conducted, identifying 22 unique studies that were included in the review.
The included studies cover a wide range of methodologies and outcomes. Supervised and unsupervised approaches are represented, including rule-based classification ( = 3), traditional Machine Learning ( = 8), Deep Learning ( = 6) and topic modelling ( = 4) methods. Datasets are mostly sourced from social media ( = 15), with other data being sourced from police forces ( = 3), health or social care providers ( = 3), or litigation texts ( = 1). Evaluation methods mostly used a held-out, labelled test set, or k-fold Cross Validation, with Accuracy and F1 metrics reported. Only a few studies commented on the ethics of computational IPV research.
Text mining methodologies offer promising data collection and analysis techniques for IPV research. Future work in this space must consider ethical implications of computational approaches.
计算文本挖掘方法被视为亲密伴侣暴力（IPV）研究中一项有用的方法创新。文本挖掘可为研究人员提供访问现有或新数据集的途径，这些数据集源自社交媒体或与IPV相关的组织，规模太大以至于无法手动分析。本文旨在概述当前在IPV研究中应用文本挖掘方法的工作，为希望在自己的研究中使用此类方法的研究人员提供一个起点。
本文报告了一项对使用计算文本挖掘研究IPV的学术研究进行系统综述的结果。根据PRISMA指南制定了综述方案，并对8个数据库进行了文献检索，确定了22项纳入综述的独特研究。
纳入的研究涵盖了广泛的方法和结果。包括监督和无监督方法，有基于规则的分类（=3）、传统机器学习（=8）、深度学习（=6）和主题建模（=4）方法。数据集大多源自社交媒体（=15），其他数据源自警察部队（=3）、卫生或社会护理提供者（=3）或诉讼文本（=1）。评估方法大多使用留出的带标签测试集或k折交叉验证，并报告了准确率和F1指标。只有少数研究评论了计算IPV研究的伦理问题。
文本挖掘方法为IPV研究提供了有前景的数据收集和分析技术。该领域未来的工作必须考虑计算方法的伦理影响。
