Article: Behavioural perception-driven evolutionary pathways for vocational English oral communication: fusing graph convolutional networks with multi-objective optimisation Journal: International Journal of Information and Communication Technology (IJICT) 2025 Vol.26 No.49 pp.75 - 91 Abstract: Existing English speaking teaching path methods ignore learners' behavioural characteristics, leading to inaccurate generation of personalised teaching paths. This paper first integrates the production-oriented approach to design the teaching process and constructs an intelligent teaching path generation model that integrates learner behavioural preference analysis and the teaching process. The model utilises graph convolutional networks and long short-term memory network to capture semantic associations and learners' dynamic evolutionary characteristics. Then, using non-dominated sorting genetic algorithm II for multidimensional optimisation, multiple paths are optimised for multiple objectives to obtain a set of frontier solutions, and the English speaking teaching path with the highest score is obtained. Experimental results show that the suggested approach improves prediction accuracy by an average of 7.04%-23.96% while significantly reducing the time required to solve for the optimal path, validating the model's efficiency. Inderscience Publishers - linking academia, business and industry through research

Title: Behavioural perception-driven evolutionary pathways for vocational English oral communication: fusing graph convolutional networks with multi-objective optimisation

Authors: Bowen Yuan; Yating Yang

Addresses: Liao Yuan Vocational Technical College, Liaoyuan, 136200, China ' Liao Yuan Vocational Technical College, Liaoyuan, 136200, China

Abstract: Existing English speaking teaching path methods ignore learners' behavioural characteristics, leading to inaccurate generation of personalised teaching paths. This paper first integrates the production-oriented approach to design the teaching process and constructs an intelligent teaching path generation model that integrates learner behavioural preference analysis and the teaching process. The model utilises graph convolutional networks and long short-term memory network to capture semantic associations and learners' dynamic evolutionary characteristics. Then, using non-dominated sorting genetic algorithm II for multidimensional optimisation, multiple paths are optimised for multiple objectives to obtain a set of frontier solutions, and the English speaking teaching path with the highest score is obtained. Experimental results show that the suggested approach improves prediction accuracy by an average of 7.04%-23.96% while significantly reducing the time required to solve for the optimal path, validating the model's efficiency.

Keywords: English speaking teaching path; production-oriented approach; graph convolutional network; long short-term memory network; NSGA-II method.

DOI: 10.1504/IJICT.2025.151060

International Journal of Information and Communication Technology, 2025 Vol.26 No.49, pp.75 - 91

Received: 09 Oct 2025
Accepted: 06 Nov 2025
Published online: 12 Jan 2026 *

Title: Behavioural perception-driven evolutionary pathways for vocational English oral communication: fusing graph convolutional networks with multi-objective optimisation

Keep up-to-date