The Enron Corpus: A New Dataset for Email Classification Research

Metadane

• Autorzy: B. Klimt, Y. Yang
• Rok: 2004
• Źródło: Machine Learning: ECML 2004, Springer, Pages 217–226
• DOI/Link: 10.1007/978-3-540-30115-8_22
• Status: to-read
• Pochodzenie: Wyekstrahowane z al-subaiey-web-ai-phishing-2024 ([35] - original Enron Corpus paper, dataset source)
• Tagi: to-read reference enron-corpus dataset-paper email-classification benchmark ecml

Notatki

Publikacja dodana automatycznie z bibliografii.

Kontekst cytowania w Al-Subaiey 2024:

• Referenced jako [35] w paper (References section)
• Original paper describing Enron Email Corpus
• Dataset used as one of 6 sources in Al-Subaiey’s merged dataset (82,486 emails total)
• Enron Corpus: foundational public email dataset enabling reproducible research

Historical Significance:

• First large-scale public email corpus (2004)
• Enabled transition from proprietary datasets to open research
• Cited thousands of times (>10,000 citations across research)
• Still widely used 20 years later (including Al-Subaiey 2024)

Dataset Characteristics (from original paper):

• ~500,000 emails from Enron Corporation employees
• Released by FERC (Federal Energy Regulatory Commission) after Enron bankruptcy
• Real-world authentic business emails (not synthetic)
• Multiple research applications: spam detection, topic classification, social network analysis

Impact on Al-Subaiey 2024:

• Provided realistic business email data
• Part of comprehensive 82k email merged dataset
• Contributes to generalizability (business context representation)

Limitation:

• Emails from 1999-2002 (dated; phishing tactics evolved)
• Al-Subaiey addresses by combining with recent datasets (CEAS, Nazario, etc.)

Dodaj PDF aby wygenerować pełne podsumowanie używając /summarize-paper klimt-enron-corpus-2004