Unveiling tortured phrases in Humanities and Social Sciences

2025-02-07

AI Summary

This study exposes the systemic risk of “tortured phrases”—convoluted, semantically opaque expressions—being weaponized by paper mills to fabricate pseudo-academic texts in the humanities and social sciences (HSS). Addressing the limitation of existing detection tools (e.g., the Paper Phrase Scanner, PPS), which are calibrated primarily for STEM domains, the work systematically identifies HSS-specific tortured phrases and constructs a novel lexicon of 121 domain-adapted abbreviation-based linguistic fingerprints. Integrating the ELSST and THESOZ social science classification schemes, the authors extend the PPS framework and apply it empirically to the EDRI journal corpus and the SSOAR repository. The analysis detects 32 interdisciplinary problematic publications—concentrated in education, psychology, and economics—validating methodological efficacy. This research fills a critical gap in automated academic misconduct detection for HSS, delivering foundational infrastructure and a scalable, reproducible methodology for scholarly integrity assessment.

Technology Category

• Application Domains: Humanities & Computational Social Science
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction

Application Category

• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

A small amount of unscrupulous people, concerned by their career prospects, resort to paper mill services to publish articles in renowned journals and conference proceedings. These include patchworks of synonymized contents using paraphrasing tools, featuring tortured phrases, increasingly polluting the scientific literature. The Problematic Paper Screener (PPS) has been developed to allow articles (re)assessment on PubPeer. Since most of the known tortured phrases are found in publications in science, technology, engineering, and mathematics (STEM), we extend this work by exploring their presence in the humanities and social sciences (HSS). To do so, we used the PPS to look for tortured abbreviations, generated from the two social science thesauri ELSST and THESOZ. We also used two case studies to find new tortured abbreviations, by screening the Hindawi EDRI journal and the GESIS SSOAR repository. We found a total of 32 multidisciplinary problematic documents, related to Education, Psychology, and Economics. We also generated 121 new fingerprints to be added to the PPS. These articles and future screening have to be investigated by social scientists, as most of it is currently done by STEM domain experts.

Problem

• Detecting tortured phrases in humanities
• Extending PPS to social sciences
• Identifying problematic multidisciplinary documents

Innovation

• Problematic Paper Screener usage
• Tortured abbreviations detection
• Multidisciplinary document screening