Article: Adaptive crawling strategies and quality-aware cleansing for Korean teaching resources Journal: International Journal of Information and Communication Technology (IJICT) 2026 Vol.27 No.69 pp.21 - 44 Abstract: This study addresses the critical challenge of automatically harvesting high-quality Korean language teaching resources from the open web, where existing methods focus on topical relevance rather than pedagogical suitability. The study proposes a novel quality-aware adaptive crawling and cleansing framework. It integrates a real-time linguistic quality assessment module, powered by universal dependencies parsing, with an adaptive crawling strategy driven by a contextual bandit algorithm. Experimental results demonstrate that quality-aware adaptive crawling and cleansing framework significantly outperforms current state-of-the-art methods. It achieves a high-quality page acquisition rate of 7.47 pages per hour (a 39% improvement), a pedagogical precision of 0.892, and a top-ranking accuracy of 0.915. The framework successfully bridges linguistic theory and web mining, offering an effective solution for building structured, high-quality pedagogical resource repositories. Inderscience Publishers - linking academia, business and industry through research

Title: Adaptive crawling strategies and quality-aware cleansing for Korean teaching resources

Authors: Jihong Quan; Jiyue Quan

Addresses: School of Foreign Languages, Liaodong University, Dandong, 118000, China ' School of business, Yulin Normal University, Yulin, 537000, China

Abstract: This study addresses the critical challenge of automatically harvesting high-quality Korean language teaching resources from the open web, where existing methods focus on topical relevance rather than pedagogical suitability. The study proposes a novel quality-aware adaptive crawling and cleansing framework. It integrates a real-time linguistic quality assessment module, powered by universal dependencies parsing, with an adaptive crawling strategy driven by a contextual bandit algorithm. Experimental results demonstrate that quality-aware adaptive crawling and cleansing framework significantly outperforms current state-of-the-art methods. It achieves a high-quality page acquisition rate of 7.47 pages per hour (a 39% improvement), a pedagogical precision of 0.892, and a top-ranking accuracy of 0.915. The framework successfully bridges linguistic theory and web mining, offering an effective solution for building structured, high-quality pedagogical resource repositories.

Keywords: adaptive web crawling; quality assessment; universal dependencies; resource cleansing.

DOI: 10.1504/IJICT.2026.154379

International Journal of Information and Communication Technology, 2026 Vol.27 No.69, pp.21 - 44

Received: 31 Dec 2025
Accepted: 02 Feb 2026
Published online: 25 Jun 2026 *

Title: Adaptive crawling strategies and quality-aware cleansing for Korean teaching resources

Keep up-to-date