Junk mail detection method and system based on dynamic update of categorizer
A spam, dynamic update technology, applied in transmission systems, digital transmission systems, electrical components, etc., can solve problems such as classification vectors that cannot save classification effects, and achieve the effect of flexible classification methods and accurate classification results.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0040] The spam dynamic detection method proposed by the invention is based on the principle of the detector and memory cells of the immune system, and the proposed spam dynamic detection system and the immune system have similarities in pattern recognition, dynamic change and noise tolerance.
[0041] Support vector machine is a classifier with statistical learning theory and excellent generalization performance. It has been successfully applied to many fields. This embodiment improves the incremental support vector machine technology in the prior art, and updates the classifier based on the principle of detectors and memory cells in the immune system.
[0042] In the updating process of the classifier by the incremental support vector machine technology in the prior art, the initial classifier is constructed by using the training samples classified as normal emails and spam emails, and the classifier includes several classification vectors representing email classification. ...
Embodiment 2
[0049] In this embodiment, a support vector machine is used to construct a classifier. The initially constructed classifier is not limited to one, but several classifiers. That is, the several classifiers can be constructed by using training samples, or one classifier can be constructed and then continuously used. The received mails construct the several classifiers, and this embodiment adopts the latter.
[0050] Since several classifiers are used in this embodiment, with the arrival of newly received emails, the classifiers with a relatively long time will be cleared, and new classifiers will be generated using newly received emails as training samples. Therefore, it can be regarded as A sliding window carries different classifiers, and the update of classifiers in this embodiment includes the following aspects:
[0051] 1) Sliding update of the classifier in the window
[0052] In this embodiment, the stream data of mails received is considered as grouped batch data, and t...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com