A Transformer-Based Semantic Control Framework for an Intelligent English Writing Assistance System with Real-Time Feedback
Abstract
English writing plays a key role in education and cross-cultural communication, yet current assistive tools struggle with coherence and adaptive feedback. This study develops an intelligent English-writing assistance system based on a Transformer-driven generative architecture enhanced by a semantic control unit and dynamic scenario attention mechanism. A large-scale corpus of about 2.1 million tokens from student essays and public datasets (ASAP, TOEFL11) was used for training and evaluation. Performance was assessed using BLEU, ROUGE-L, BERTScore, METEOR, and TaskFit indices under consistent hardware settings. In a comparative experiment with 30 participants, average writing time was reduced by 28 %, revision count by 22 %, and recommendation adoption rate rose to 78.3 %. Usability improved significantly (SUS 68.5→84.2) while workload decreased (NASA-TLX –15 points). These results demonstrate that the proposed system enhances logical coherence, semantic accuracy, and user experience through an integrated feedback loop architecture linking generation, evaluation, and error-driven optimization.References
Yuan L .Research on English Hybrid Assisted Teaching System Using Contextual Support of R-CNN[J].Wireless Communications & Mobile Computing, 2022.https://dol:10.1155/2022/5358546.
Mei Z , Lin Z .Design of Vocabulary Query System in Computer Aided English Translation Teaching[C]//EAI International Conference, BigIoT-EDU.Springer, Cham, 2024.https://dol:10.1007/978-3-031-63136-8_8.
Nikolopoulos A , Karalis V D .Implementation of a Generative AI Algorithm for Virtually Increasing the Sample Size of Clinical Studies[J].Applied Sciences (2076-3417), 2024, 14(11).https://dol:10.3390/app14114570.
Chai J .On Teaching Graduate English Academic Writing Based on Artificial Intelligence[J].Procedia Computer Science, 2024, 247(000):1366-1373.https://dol:10.1016/j.procs.2024.10.163.
Wang C , Li Z , Bonk C .Understanding self-directed learning in AI-Assisted writing: A mixed methods study of postsecondary learners[J].Computers and Education: Artificial Intelligence, 2024, 6.https://dol:10.1016/j.caeai.2024.100247.
Neunzig C , Mllensiep D , Kuhlenktter B ,et al.ML Pro: digital assistance system for interactive machine learning in production[J].Journal of Intelligent Manufacturing, 2024, 35(7).https://dol:10.1007/s10845-023-02214-0.
Jingning L .Speech recognition based on mobile sensor networks application in English education intelligent assisted learning system[J].Measurement: Sensors, 2024, 32(000).https://dol:10.1016/j.measen.2024.101084.
Jiang L , Wang Y .English Online Teaching Assistant System Based on Genetic Algorithms[C]//EAI International Conference, BigIoT-EDU.Springer, Cham, 2024.https://dol:10.1007/978-3-031-63136-8_49.
Tu X , Zhang H .Educational Service Platform and English Corpus Design Based on Embedded Intelligent System and Cloud Computing[J].Wireless Communications & Mobile Computing, 2022.https://dol:10.1155/2022/1916941.
Dai B .Simulation of interactive E-learning mode based on association rule mining in computer English reading assistance system[J].Entertainment Computing, 2024, 51.https://dol:10.1016/j.entcom.2024.100742.
Zhao Q , Nazir S .English Multimode Production and Usage by Artificial Intelligence and Online Reading for Sustaining Effectiveness[J].Mobile Information Systems, 2022.https://dol:10.1155/2022/6780502.
Pack A , Barrett A , Escalante J .Large language models and automated essay scoring of English language learner writing: Insights into validity and reliability[J].Computers and Education: Artificial Intelligence, 2024, 6.https://dol:10.1016/j.caeai.2024.100234.
Lee Y J , Davis R O .A case study of implementing generative AI in university's general English courses[J].Contemporary Educational Technology, 2024, 16(4).https://dol:10.30935/cedtech/15218.
Yang S H .Enhancing English as a Foreign Language Learners' Writing Performance and Experience With an Artificial Intelligence Powered Tool[J].Springer, Cham, 2024.https://dol:10.1007/978-3-031-76293-2_9.
Liu M , Zhang L J , Biebricher C .Investigating students' cognitive processes in generative AI-assisted digital multimodal composing and traditional writing[J].Computers & education, 2024, 211(Apr.):1.1-1.21.https://dol:10.1016/j.compedu.2023.104977.
Han X .An Improved Classification Model for English Syntax Error Correction Design of DL Algorithm[J].International Arab Journal of Information Technology (IAJIT), 2024, 21(4).https://dol:10.34028/iajit/21/4/2.
Gustilo L , Ong E , Lapinid M R .Algorithmically-driven writing and academic integrity: exploring educators' practices, perceptions, and policies in AI era[J].International Journal for Educational Integrity, 2024, 20(1):43.https://dol:10.1007/s40979-024-00153-8.
Asad M M , Shahzad S , Shah S H A ,et al.ChatGPT as artificial intelligence-based generative multimedia for English writing pedagogy: challenges and opportunities from an educator's perspective[J].International Journal of Information and Learning Technology, 2024, 41(5).https://dol:10.1108/IJILT-02-2024-0021.
Ali S .Locating the Intersection of Generative Artificial Intelligence and Human English Writing Skills: A Comparative Study[J].Arab World English Journal, 2024.https://dol:10.24093/awej/ChatGPT.7.
Zenni R D , Andrew N R .Artificial Intelligence text generators for overcoming language barriers in ecological research communication[J].Austral Ecology, 2023(7):48.https://dol:10.1111/aec.13417.
DOI:
https://doi.org/10.31449/inf.v49i20.10846Downloads
Published
How to Cite
Issue
Section
License
Authors retain copyright in their work. By submitting to and publishing with Informatica, authors grant the publisher (Slovene Society Informatika) the non-exclusive right to publish, reproduce, and distribute the article and to identify itself as the original publisher.
All articles are published under the Creative Commons Attribution license CC BY 3.0. Under this license, others may share and adapt the work for any purpose, provided appropriate credit is given and changes (if any) are indicated.
Authors may deposit and share the submitted version, accepted manuscript, and published version, provided the original publication in Informatica is properly cited.
