EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Fragment-Level Evidence Retrieval for Complex Long Texts in the Power Domain via Recursive Character Splitting

Xiawei Cheng, Wanden Mao, Joshua Luhwago, Asad Ullah, Guojing Qian, Shengzhe Xi, Shanfeng Liu, Li Ruan, Limin Xiao

Electronics · 18.11.2026 · 15 (22) · 5373

DOI: 10.3390/electronics15225373 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Electronics, online erschienen am 18.11.2026. Verfasst von Xiawei Cheng und 8 weiteren Autor:innen. Die Arbeit nennt 17 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: State Grid Henan Electric Power Company.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Knowledge bases for primary power equipment must retrieve concise, verifiable evidence from long standards, fault cases, and maintenance records. We evaluate six retrieval-unit strategies, three retrievers, and two Chinese embedding encoders using 149 documents (1.73 million characters) and 48 author-constructed questions, with 20 for development and 28 held out for testing. At the matched 500-character/100-overlap setting, recursive splitting preserves 84.5% of numbered standards clauses, compared with 78.6% for sliding-window, 72.0% for paragraph, and 64.9% for semantic chunking; clause-aware splitting reaches 88.3%. Recursive splitting with dense retrieval records Hit@10 of 0.571 and evidence coverage@10 of 0.542; hybrid dense–BM25 fusion raises these to 0.714 and 0.685. Sliding-window dense retrieval records 0.786 and 0.685 on the same measures. Recursive, sliding-window, paragraph, and semantic chunks each contain all 112 annotated evidence spans, whose lengths range from 11 to 90 characters (median 36); clause-aware chunking contains 97.3%. The results characterize the structural-preservation and retrieval profiles of the tested strategies and show that hybrid fusion improves top-10 evidence recovery with recursive chunks.

Förderung

  • State Grid Henan Electric Power Company

Zitieren

APA

Cheng, X., Mao, W., Luhwago, J., Ullah, A., Qian, G., Xi, S., Liu, S., Ruan, L., & Xiao, L. (2026). Fragment-Level Evidence Retrieval for Complex Long Texts in the Power Domain via Recursive Character Splitting. Electronics, 15(22), 5373. https://doi.org/10.3390/electronics15225373

Vancouver

Cheng X, Mao W, Luhwago J, Ullah A, Qian G, Xi S, et al. Fragment-Level Evidence Retrieval for Complex Long Texts in the Power Domain via Recursive Character Splitting. Electronics. 2026;15(22):5373. doi:10.3390/electronics15225373

BibTeX
@article{Cheng2026,
  author = {Cheng, Xiawei and Mao, Wanden and Luhwago, Joshua and Ullah, Asad and Qian, Guojing and Xi, Shengzhe and Liu, Shanfeng and Ruan, Li and Xiao, Limin},
  title = {Fragment-Level Evidence Retrieval for Complex Long Texts in the Power Domain via Recursive Character Splitting},
  journal = {Electronics},
  year = {2026},
  volume = {15},
  number = {22},
  pages = {5373},
  doi = {10.3390/electronics15225373}
}
Kostenlos starten →