Author

Xiaojun Yuan

View on Semantic Scholar

2 papers indexed here

We haven’t gathered this author’s papers yet. Follow them and we’ll fetch their work.

Not the right person? Other researchers publish under this name.

Review Open access Jul 2026

AI-Driven Feedback for University Students’ ESL/EFL Academic Writing: A Scoping Review of Pedagogical Implementation and Learning Outcomes

This review maps the recent findings on the AI-driven feedback given to undergraduate-level ESL/EFL academic writing, examining how it is implemented, what learning outcomes it produces, and what shapes those outcomes. Drawing on 39 peer-reviewed empirical studies retrieved from Web of Science, Scopus, and ERIC, and following the PRISMA extension for Scoping Reviews, this review covers 2,796 participants across 18 countries and regions. The findings show that 27 of the 39 studies in the full corpus reported measurable improvement in ESL/EFL writing. Lower-proficiency students showed larger surface-level gains but higher rates of passive uptake, while higher-proficiency learners engaged more critically when appropriately scaffolded. This review also identifies a persistent local-global revision gap: grammar and vocabulary improvements were reliably documented, but gains in argumentation, genre awareness, and critical reasoning remained weak and variable across the corpus. While AI-driven feedback shows genuine promise for addressing the feedback deficit in large-scale ESL/EFL writing instruction — a challenge with direct relevance to SDG 4 (Quality Education) and its advocation for inclusive, equitable access to quality learning at all levels - the conditions that produce durable learning gains remain poorly understood. The field's methodological development has not yet caught up with its empirical output.

Xiaojun Yuan, Nee Nee Chan · International Journal of Lea... · 0 citations

Preprint Jul 2026

H+ Embedding: Harmonizing Global and Token-Level Retrieval with Context-Dependent Phrases

Terminology-intensive retrieval, especially in medical settings, depends on preserving multi-word entities, abbreviations, numerical constraints, and compositional concepts. However, existing representations lie at two extremes: single-vector retrievers often over-compress local relevance signals, while token-level late interaction retains every tokenizer subword at substantial indexing, storage, and scoring cost. This mismatch raises a natural question: can context-dependent phrases provide a useful retrieval unit between global vectors and tokens? We introduce H+ Embedding, a unified multi-granularity retriever that predicts variable-length phrase partitions, preserves uncovered tokens as singletons, and applies importance-guided unit selection with weighted MaxSim interaction. Across 16 scientific, medical, and bilingual tasks, its phrase retrieval branch exceeds the global retrieval branch by 6.91 macro nDCG@10. It also nearly matches Token while using 13.7% fewer document vectors and outperforms content-independent grouping rules under moderate vector budgets. Context-dependent phrase interaction therefore provides an intermediate quality-cost point between global compression and token-level interaction for practical retrieval systems.

Shusen Zhang, Junyi Hu, Ye Feng et al. · 0 citations