Jul 2026· Journal of Intelligent Decision Making and Information Science· Vol 3, pp. 1128-1141· 0 citations· 24 references
TL;DR
A statistically significant increase in lexical density and frequency of cohesive markers has been revealed, indicating an increase in information compression and explicit discursive organization of texts, consistent with characteristics described in AI-assisted writing studies.
Abstract
The rapid spread of large language models (LLMs) has significantly transformed academic writing practices and actualized discussions about authorship, language quality, and academic integrity. At the same time, diachronic changes in academic discourse during the active implementation of generative artificial intelligence remain insufficiently studied. The study combines a systematic literature review with a corpus diachronic analysis of authentic academic annotations, allowing us to trace long-term trends in the development of academic discourse. The aim of the work is to identify linguistic changes in academic writing during 2015–2025 and determine the role of human editing in quality assurance of AI-assisted scientific texts. The research material was a corpus of 870 English-language annotations of scientific articles indexed in the Scopus database in the field of arts and humanities. Quantitative linguistic analysis was carried out using Python tools. Indicators of lexical density, lexical diversity, syntactic complexity, average sentence length and frequency of cohesive markers were analyzed. A statistically significant increase in lexical density and frequency of cohesive markers has been revealed, indicating an increase in information compression and explicit discursive organization of texts. Indicators of traditional lexical diversity and syntactic complexity remained relatively stable. The observed trends are consistent with characteristics described in AI-assisted writing studies; however, the study design does not allow them to be directly related to the use of large language models. Human editing remains a key factor in ensuring factual accuracy, discursive coherence, lexical enrichment, and academic integrity. The results can be used to develop practices for the responsible use of generative AI in academic communication.
The results revealed relevant differences between the analyzed papers: texts produced in 2023 showed greater stylistic variation, the presence of authorial markers, and irregularities typical of human writing, whereas texts from 2025 presented a higher concentration of indicators associated with linguistic standardization, structural uniformity, and a reduction of individual markers of authorship.
Gabriela Pereira da Silva, I. Rhuan, G. Tardo et al.· 0 citations
The rapid expansion of generative artificial intelligence and large language models has dramatically altered the
creation of written material. While these technologies can enhance the learning process, but, as well, they have
created significant concerns regarding academic integrity, authorship and the fair assessment of students. In response,
numerous institutions have already implemented AI-detection software that aims at determining whether some text was
authored by a human or generated by AI systems. But there are concerns about the validity, correctness and equity of
such tools in the actual fields of education. This paper gives a systematic review of empirical studies concerning AI
text detection tools in learning institutions. In order to identify literature published between 2022 and 2026, it was
thoroughly searched in large academic databases such as Scopus, Web of Science, and ScienceDirect, following the
PRISMA 2020 guidelines. Through the use of rigorous inclusion/exclusion criteria, a total of 17 empirical studies were
selected for in-depth analysis. In the review, the researchers examine the different types of research methodologies used
to evaluate the performance of AI detection tools, the accuracy of these tools in identifying AI-generated content, and
the effectiveness, fairness, and usability of these tools. The results show that the tools, including Turnitin, GPTZero,
Copyleaks, ZeroGPT, and Originality.ai, are frequently used, but their effectiveness varies depending on the type of
text and writing conditions. It is often mentioned in the literature that such systems become quite easily affected by
paraphrasing, translation or writing style, and that they can falsely categorize the texts written by multilingual or nonnative writers of English. The review also presents a number of technical, ethical and pedagogical difficulties with the
continuation of the use of automated detection systems only. Upon the evidence synthesized, the study elucidates the
importance of careful and open usage of AI detection technologies in learning and suggests the directions of future
research to enhance its consistency and at the same time encourages ethical academic procedures.
Arslan Akram· Advances in Machine Learning...· 0 citations
The findings show that AI-generated texts exhibit greater lexical diversity and syntactic complexity; however, they often exhibit structural uniformity, overuse of cohesive devices, and limited pragmatic depth, and should not replace professionally designed educational materials.
V. Smaglii, T. Korolova, Svitlana Yukhymets et al.· Arab World English Journal· 0 citations
Analysis of AIGC texts points out that the complexity of AI text primarily stems from its mechanism of selecting vocabulary based on probability distributions, which favors longer words, abstract nouns, and words with high semantic content, thereby forming a highly compact linguistic surface.
Lulu Chen· Lecture Notes on Language an...· 0 citations
This volume contains abstracts of conference papers by the university professors in the field of applied and theoretical translation studies and linguistics, as well as by translators and interlingual mediators from the European Parliament. The contributions in English and German address the use of artificial intelligence in various fields of language-related professional practice, particularly in translation, proofreading and editing, terminology management, language consulting, and linguistic research. The authors present contemporary tools based on large language models and discuss their potential for automating routine tasks, analysing textual data, and supporting the production of high-quality language content. The contributions also highlight issues of reliability, terminological consistency, ethical responsibility, and data protection in the use of artificial intelligence within the process of interlingual transfer. Artificial intelligence is undoubtedly becoming an important supportive element of interlingual mediation; however, its effective and responsible use requires the active involvement, critical judgement, and a high level of professional expertise on the part of language professionals.
Unknown authors· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
