基于语料库的人机学术英语摘要语言特征多维对比研究
A Multidimensional Comparative Study on Linguistic Features of Human-Machine Academic English Abstracts Based on Corpus
DOI: 10.12677/ml.2026.14111042, PDF,    科研立项经费支持
作者: 解佳颖：西安电子科技大学语言智能学部，陕西 西安
关键词: 语料库；大语言模型；学术英语摘要；句法复杂度；随机森林；Corpus； Large Language Model； Academic English Abstract； Syntactic Complexity； Random Forest
摘要: 为探究人机学术写作差异，本研究构建平行语料库，提取多维计量指标，运用随机森林算法对比人类专家与大模型的英文摘要。结果表明，大模型虽能拟合宏观连贯，但在微观语用策略上模型倾向堆砌复杂词汇与过度嵌套或扁平化的极端句式以刻意模拟学术感，难以像人类般自然平衡文本信息。研究证实深层句法与词汇特征能甄别人机文体英文摘要，为揭示机器生成机制与优化语篇甄别提供了实证支持。
Abstract: To explore the discrepancies between human and AI academic writing, this study constructs a parallel corpus and extracts multi-dimensional quantitative metrics, and adopts the random forest algorithm to compare English abstracts written by human scholars and large language models. The results reveal that although large language models can reproduce macro textual coherence, they adopt biased micro-pragmatic strategies: they tend to overuse sophisticated vocabulary and employ extreme sentence structures featuring excessive embedding or flat syntax to artificially simulate an academic tone, and fail to balance textual information naturally as human writers do. This study verifies that deep syntactic and lexical features can distinguish human-written abstracts from AI-generated ones, and provides empirical evidence for uncovering the generation mechanism of large language models and improving the discrimination of machine-produced texts.
文章引用：解佳颖. 基于语料库的人机学术英语摘要语言特征多维对比研究[J]. 现代语言学, 2026, 14(11): 19-28. https://doi.org/10.12677/ml.2026.14111042

参考文献

[1] Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J.D., Dhariwal, P. and Amodei, D. (2020) Language Models Are Few-Shot Learners. Advances in Neural Information Processing Systems 33: Annual Conference on Neural Information Processing Systems 2020, NeurIPS 2020, 6-12 December 2020, 1877-1901.
[2] Fredrick, D.R. and Craven, L. (2025) Lexical Diversity, Syntactic Complexity, and Readability: A Corpus-Based Analysis of ChatGPT and L2 Student Essays. Frontiers in Education, 10, Article 1616935.
https://doi.org/10.3389/feduc.2025.1616935
[3] Shen, C., Guo, J., Shi, P., Qu, S. and Tian, J. (2023) A Corpus-Based Comparison of Syntactic Complexity in Academic Writing of L1 and L2 English Students across Years and Disciplines. PLOS ONE, 18, e0292688.
https://doi.org/10.1371/journal.pone.0292688
[4] Biber, D. and Gray, B. (2016) Grammatical Complexity in Academic English: Linguistic Change in Writing. Cambridge University Press.
https://doi.org/10.1017/cbo9780511920776
[5] Kyle, K. and Crossley, S. (2017) Assessing Syntactic Sophistication in L2 Writing: A Usage-Based Approach. Language Testing, 34, 513-535.
https://doi.org/10.1177/0265532217712554
[6] McNamara, D.S. and Graesser, A.C. (2012) Coh-Metrix: An Automated Tool for Theoretical and Applied Natural Language Processing. In: McCarthy, P.M. and Boonthum-Denecke, C., Eds., Applied Natural Language Processing: Identification, Investigation and Resolution, IGI Global, 188-205.
https://doi.org/10.4018/978-1-60960-741-8.ch011
[7] 吴继峰, 胡韧奋, 彭一平. 大语言模型和汉语母语者生成学术汉语语言特征对比研究[J]. 语言文字应用, 2025(4): 27-37.
[8] Casal, J.E. and Kessler, M. (2023) Can Linguists Distinguish between ChatGPT/AI and Human Writing? A Study of Research Ethics and Academic Publishing. Research Methods in Applied Linguistics, 2, Article ID: 100068.
https://doi.org/10.1016/j.rmal.2023.100068
[9] 张鹏华, 潘一. 中外学位论文英文摘要语步对比研究[J]. 现代语言学, 2023, 11(5): 2181-2192.
[10] 卜小伟, 杨云峰. 英汉学术论文摘要语篇衔接的对比研究[J]. 长沙铁道学院学报: 社会科学版, 2010, 11(3): 171-172.
[11] 朱君辉, 王梦焰, 杨尔弘, 等. 大模型生成回答与人类回答文本的语言特征比较研究[J]. 中文信息学报, 2024, 38(4): 17-27.
[12] Solaiman, I., Brundage, M., Clark, J., Askell, A., Herbert-Voss, A., Wu, J. and Wang, J. (2019) Release Strategies and the Social Impacts of Language Models. arXiv: 1908.09203.
[13] Uchendu, A., Le, T., Shu, K. and Lee, D. (2020) Authorship Attribution for Neural Text Generation. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), 16-20 November 2020, 8384-8395.
https://doi.org/10.18653/v1/2020.emnlp-main.673
[14] 崔中良. 生成式人工智能是否能替代人类生成语言?——从梅洛-庞蒂的语言生成观探讨[J]. 西安交通大学学报 (社会科学版), 2025, 45(3): 182-190.
[15] Moslem, Y., Haque, R., Kelleher, J. and Way, A. (2023) Adaptive Machine Translation with Large Language Models. Proceedings of the 24th Annual Conference of the European Association for Machine Translation, Tampere, 12-15 June 2023, 227-237.
[16] Kwok, H.L., Shi, Y., Xu, H., Li, D. and Liu, K. (2025) GenAI as a Translation Assistant? A Corpus-Based Study on Lexical and Syntactic Complexity of GPT-Post-Edited Learner Translation. System, 130, Article ID: 103618.
https://doi.org/10.1016/j.system.2025.103618
[17] Swain, M. (1993) The Output Hypothesis: Its History and Its Future. In: Cook, G. and Seidlhofer, B., Eds., Principle and Practice in Applied Linguistics: Studies in Honour of H.G. Widdowson, Oxford University Press, 125-144.
[18] Breiman, L. (2001) Random Forests. Machine Learning, 45, 5-32.
https://doi.org/10.1023/a:1010933404324
[19] Kuhn, M. (2008) Building Predictive Models in R Using the Caret Package. Journal of Statistical Software, 28, 1-26.
https://doi.org/10.18637/jss.v028.i05