Fraud website identification method and system based on picture clustering and artificial study and judgment
A recognition method and image technology, applied in character and pattern recognition, website content management, network data retrieval, etc., can solve problems such as inability to identify new types of fraudulent websites, lack of discovery of new fraudulent websites, etc., to improve representativeness and reliability , Improve recognition ability, improve the effect of discovery ability
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0050] 1. Extract feature vectors and hash feature values from known fraud-related websites, and enter them into the fraud-related sample database.
[0051] 2. Calculate the Euclidean distance on the feature vectors of the fraud sample database in a recursive manner, and perform clustering based on the Euclidean distance.
[0052] 3. For a batch of unknown websites, the image feature vectors of the websites are obtained through the feature model.
[0053] 4. Calculate the Euclidean distance between the image feature vector and each type of cluster center feature vector of the fraudulent sample library, and judge the type of the unknown website by the distance. If the distance is 0, it is directly judged that the website is an illegal website. For those whose distance is not 0, it is regarded as a suspected illegal website, and the matching type is regarded as a suspected type.
[0054] 5. Manually analyze the suspected types obtained from this batch of unknown websites, and...
Embodiment 2
[0059] Such as figure 1 As shown, a fraudulent website identification method based on image clustering and manual judgment is provided, including:
[0060] S1: Obtain effective website screenshots of existing fraudulent websites and their fraudulent types, perform feature extraction on fraudulent images, obtain image fusion features as fraud samples, and classify according to the fraudulent types of image fusion features; S1 is specifically:
[0061] S1.1. Obtain existing known fraudulent websites and their types.
[0062] S1.2. Build a model based on the mobilenetv3 network and pre-training weights based on the imageNet dataset.
[0063] S1.3. Based on the constructed pre-training model combined with the existing data of fraudulent websites, transfer learning is performed to learn the feature distribution of fraudulent websites.
[0064] S1.4. Use a transfer learning model that fits the distribution of fraudulent websites as a feature extractor.
[0065] S1.5. Obtain the g...
PUM
![No PUM](https://static-eureka.patsnap.com/ssr/23.2.0/_nuxt/noPUMSmall.5c5f49c7.png)
Abstract
Description
Claims
Application Information
![application no application](https://static-eureka.patsnap.com/ssr/23.2.0/_nuxt/application.06fe782c.png)
- 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.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap