Email Leakage Detection via Non-Sensitive Attribute Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing information leakage detection systems in emails rely on content inspection, which is resource-intensive, slow, and may expose sensitive information, and do not effectively utilize sender and recipient profiles to identify anomalies.
Innovation Solution
A system that extracts non-sensitive attributes from email headers and sender profiles using filtering techniques, employing neural networks and support vector machines to detect information leakage without analyzing email content, thus reducing hardware requirements and improving speed and reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If content inspection is used to detect information leakage, then detection accuracy is improved, but system complexity and resource requirements increase
Solution Approach 1:
The patent extracts only non-sensitive attributes from email content such as sender profile information, recipient information, time of sending, and frequency of mails. By taking out only the necessary non-sensitive features rather than analyzing full content, the system achieves detection capability while reducing complexity and resource requirements.
Solution Approach 2:
The patent segments the email analysis into distinct non-sensitive attribute components (sender profile, recipient info, timing, frequency) that can be independently extracted and analyzed. This segmentation allows the system to focus on specific detection-relevant features without processing the entire email content, reducing overall system complexity.
2Measurement precision
If full email content scanning is performed, then information leakage detection capability is improved, but processing speed decreases
Solution Approach 1:
The system extracts only essential non-sensitive attributes (sender profile, recipient info, time, frequency) from emails rather than scanning full content. This extraction approach maintains detection capability for information leakage patterns while significantly reducing processing time and improving throughput.
Solution Approach 2:
The patent applies partial action by analyzing only the necessary non-sensitive portion of emails (attributes like sender profile, timing, frequency) rather than performing excessive full-content scanning. This partial analysis approach achieves sufficient detection accuracy while maintaining high processing speed.
3Reliability
If email content is scanned for detection, then detection reliability is improved, but user privacy protection deteriorates
Solution Approach 1:
The patent extracts only non-sensitive attributes from emails such as sender profile, recipient information, sending time, and frequency. By taking out only these non-sensitive features and excluding sensitive content, the system maintains detection reliability while protecting user privacy and preventing exposure of confidential information.
Solution Approach 2:
Instead of the conventional approach of scanning full email content to ensure detection reliability, the patent inverts the approach by deliberately excluding sensitive content and relying solely on non-sensitive attributes. This inversion maintains detection capability through pattern analysis of non-sensitive features while inherently protecting privacy.
Data Source
AI summary
A system for detecting information leakage in e-mails using neural network and support vector machines is provided. This system does not use the content of the e-mail or the content of the attachments in the e-mail. Instead, a set of non-sensitive variables or attributes is picked from the e-mails originating from a given establishment and also from the profiles of the users sending those mails. The said attributes are extracted for all outbound mails. This extraction process does not involve reading the main text of the mail and thus the sensitivity of the mail information is protected. These attributes are chosen using filters built into the detection hardware. Neural networks and support vector machine built into the detection hardware are then used on these attributes to detect pattern violation and possible information leakage.