Analytics & Metrics Overview

71,000 Records

Statistical distributions, risk curves, keyword density, and sender domain breakdown

Dataset Scale

71,000

Based on: All records in data.csv

Phishing Ratio

55.8%

Column: label (1 vs 0)

History

Logged in: history.json

Blocked Domains

Registry: blacklist.json

Analyzed Dataset Columns (8 Fields)

The core data structure and attributes evaluated across our phishing analysis pipeline

CSV Schema

labelClassification Target

Type: Integer (0 | 1)

Ground truth classification: 1 indicates Phishing, 0 indicates Safe/Legitimate.

senderOrigin Header

Type: Text String

Sender email address and display name used to trace origins and extract sender domains.

subjectHeader Content

Type: Text String

Email subject line analyzed for urgency triggers, threat cues, and capitalization patterns.

bodyPayload Content

Type: Multiline Text

Full message text inspected for suspicious keywords, deceptive phrases, and call-to-actions.

domainExtracted Entity

Type: Hostname

Extracted root domain cross-referenced against the active blacklist and high-risk domain registries.

urlsExtracted Links

Type: Array of Links

Collection of external web links and direct IP addresses detected inside the email body.

receiverRecipient Header

Type: Text String

Target recipient address or destination mailbox associated with the email record.

dateTemporal Data

Type: Date / Timestamp

Chronological timestamp indicating when the email message was originally logged.

Statistical Risk Score Metrics

Based on: label, subject, body

Distribution parameters calculated across evaluated records based on the label classification, with risk weights from subject and body.

Sample size: 1000

Mean Score—Average Score

Std Deviation—Spread & Variance

Median—50th Percentile

25th Percentile—Lower Quartile

75th Percentile—Upper Quartile

Score Range0–100Min to Max

Data Source Attribution: Metrics computed across evaluated dataset records based on label (ground truth), subject (header indicators), and body (content risk factors).

Risk Score Distribution Histogram

Based on: label, body, subject

Frequency distribution of evaluated emails grouped into 10-point risk score intervals

Safe (0–39) Suspicious (40–69) Phishing (70–100)

Most Frequent Phishing Keywords

Based on: body, subject

Keywords detected in body and subject of phishing emails (label = 1).

Phishing Sender Domain Breakdown

Based on: sender, domain, label

Origin domains extracted from sender column for records flagged as phishing (label = 1).

Domain	Phishing Volume