EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Rule-Based 5W1H Information Extraction for Phishing Email Analysis in Digital Forensic Investigations

Aura Fadhila Kusumaningtyas, Erika Ramadhani

bit-Tech · 10.08.2026 · 9 (1) · 1105-1114

DOI: 10.32877/bt.v9i1.4053 ↗

Kurz gefasst

Fachartikel in der Zeitschrift bit-Tech, online erschienen am 10.08.2026. Verfasst von Aura Fadhila Kusumaningtyas und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Phishing email evidence in digital forensic investigations requires investigators to extract structured information covering sender identity, requested actions, timing, attack infrastructure, motivation, and method, yet this process remains largely manual. Six participants with digital forensic investigation experience reported that documentation of a single case routinely takes more than five hours, that important evidence is occasionally missed despite repeated manual review, and that any automated extraction tool must remain fully explainable to support forensic documentation requirements. This study develops a rule-based 5W1H information extraction system for phishing email analysis using the full Design Thinking methodology. Empathize and Define stages established thirteen system requirements, prioritizing extraction comprehensiveness and auditable regex-based logic over opaque machine learning approaches. The resulting system applies regex-based pattern matching across eight extraction categories spanning six 5W1H dimensions, implemented through a Flask backend and web-based frontend. Technical evaluation against fifty manually annotated emails from the CEAS_08 dataset produced an overall F1-score of 81.19%, with WHERE (URLs and Domains) reaching 96.02% and WHEN (Date and Urgency) reaching 66.29%, the latter driven by a recurring double-extraction pattern in timestamp fields. Usability evaluation with the same six participants produced an overall satisfaction rating of 4.17 out of 5. The evaluation is preliminary, drawing on fifty emails from a single dataset and usability input from the same six participants. Nonetheless, the results suggest that rule-based regex extraction offers a technically viable and auditable approach to structured 5W1H documentation support in phishing email forensic investigation, suitable for further evaluation in operational forensic settings.

Zitieren

APA

Kusumaningtyas, A. F., & Ramadhani, E. (2026). Rule-Based 5W1H Information Extraction for Phishing Email Analysis in Digital Forensic Investigations. bit-Tech, 9(1), 1105-1114. https://doi.org/10.32877/bt.v9i1.4053

Vancouver

Kusumaningtyas AF, Ramadhani E. Rule-Based 5W1H Information Extraction for Phishing Email Analysis in Digital Forensic Investigations. bit-Tech. 2026;9(1):1105-1114. doi:10.32877/bt.v9i1.4053

BibTeX
@article{Kusumaningtyas2026,
  author = {Kusumaningtyas, Aura Fadhila and Ramadhani, Erika},
  title = {Rule-Based 5W1H Information Extraction for Phishing Email Analysis in Digital Forensic Investigations},
  journal = {bit-Tech},
  year = {2026},
  volume = {9},
  number = {1},
  pages = {1105--1114},
  doi = {10.32877/bt.v9i1.4053}
}
Kostenlos starten →