Magnitude Score Weights

Phinder calculates a Magnitude Score for each file and an aggregate score for the entire scan. By default, every PII occurrence has a weight of 1.0.

Note

The Magnitude Score measures how much PII a file holds, not how sensitive it is. To rank files by sensitivity instead, see Risk Scoring, which is calculated separately and configured separately. Weights tune the Magnitude Score only and do not affect the Risk Score.

Custom Weights

You can specify custom weights for different PII types using a weights.json file. This allows you to prioritize certain types of PII over others (e.g., a SSN might have a higher magnitude than an email address).

weights.json Format

The file should be a JSON object mapping PII types (as identified by Phileas) to numeric weights.

{
  "email-address": 2.0,
  "ssn": 10.0,
  "phone-number": 5.0
}

Using Weights

Specify the weights file with the -w or --weights option:

java -jar phinder.jar -i documents/ -w my-weights.json

Calculation

The Magnitude Score is calculated as: Sum(PII Type Count * PII Type Weight)

Density Score

Phinder also calculates a Density Score, which is the Magnitude Score divided by the word count of the document: Density Score = Magnitude Score / Word Count

This score helps identify documents with a high concentration of PII relative to their size.

Weights and Severities

Weights and severities look alike and are easy to confuse:

	Weight	Severity
Feeds	Magnitude Score, Density Score	Risk Score
Default	1.0 for every type	Per type, from the built-in severity table
Option	--weights	--severities

Setting one has no effect on the other, so you can tune each for what it measures.