Article: Interactive mobile English translation teaching quality evaluation model based on particle swarm optimisation-neural network Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.97 pp.1 - 20 Abstract: In the context of the increasing importance of interactive mobile English translation teaching and aiming at the real-time and accuracy challenges of quality evaluation, this paper proposes an innovative model based on particle swarm optimisation-neural network (PSO-NN). The model dynamically adjusts the hyperparameters of the LSTM-attention network through adaptive multi-scale particle swarm optimisation (AMPSO), combines attention mechanism to enhance time series feature extraction, and adopts a unified preprocessing process to process multi-source data. The experimental verification results show that the F1-score of the model reaches 0.895, the accuracy rate is 89.5%, and the AUC is 0.951 in the performance comparison, which is significantly better than the baseline. In the robustness test, ΔF1 of the model only decreases by 4.1% under noisy data. The practical test shows that the inference time is 50.1 ms, the memory occupation is 130.5 MB, and the efficiency and real-time performance are considered. This model effectively improves the evaluation accuracy and anti-interference ability, and its innovation lies in the collaborative optimisation of AMPSO's multi-scale mutation strategy and neural network components, which provides a reliable tool for mobile teaching. Inderscience Publishers - linking academia, business and industry through research

Title: Interactive mobile English translation teaching quality evaluation model based on particle swarm optimisation-neural network

Authors: Jing Tu

Addresses: The School of Foreign Languages, Nanchang Institute of Technology, 330044, Jiangxi, China

Abstract: In the context of the increasing importance of interactive mobile English translation teaching and aiming at the real-time and accuracy challenges of quality evaluation, this paper proposes an innovative model based on particle swarm optimisation-neural network (PSO-NN). The model dynamically adjusts the hyperparameters of the LSTM-attention network through adaptive multi-scale particle swarm optimisation (AMPSO), combines attention mechanism to enhance time series feature extraction, and adopts a unified preprocessing process to process multi-source data. The experimental verification results show that the F1-score of the model reaches 0.895, the accuracy rate is 89.5%, and the AUC is 0.951 in the performance comparison, which is significantly better than the baseline. In the robustness test, ΔF1 of the model only decreases by 4.1% under noisy data. The practical test shows that the inference time is 50.1 ms, the memory occupation is 130.5 MB, and the efficiency and real-time performance are considered. This model effectively improves the evaluation accuracy and anti-interference ability, and its innovation lies in the collaborative optimisation of AMPSO's multi-scale mutation strategy and neural network components, which provides a reliable tool for mobile teaching.

Keywords: particle swarm optimisation-neural network; PSO-NN; interactive; English translation; quality of teaching; evaluation model.

DOI: 10.1504/IJICT.2026.156200

International Journal of Information and Communication Technology, 2026 Vol.27 No.97, pp.1 - 20

Received: 07 Jan 2026
Accepted: 06 Mar 2026
Published online: 07 Sep 2026 *

Title: Interactive mobile English translation teaching quality evaluation model based on particle swarm optimisation-neural network

Keep up-to-date