Web Page Classification Algorithm Based on Deep Learning.

Area of Science:

• Computer Science
• Artificial Intelligence

Background:

• Current research on Web Page Classification Algorithms (WPCA) using Deep Learning (DL) lacks depth.
• Efficient processing of image and sound data requires robust learning mechanisms.

Purpose of the Study:

• To research and develop an improved WPCA utilizing Deep Learning (DL).
• To enhance the accuracy and efficiency of web page classification.
• To optimize the learning rate for DL models.

Main Methods:

• Implemented a keyword weight calculation method to refine word importance.
• Utilized a similarity-based classification approach for Chinese web pages.
• Employed adaptive parameters and optimization algorithms to adjust DL learning rates.

Main Results:

• The DL-based WPCA demonstrated superior performance compared to traditional algorithms.
• Time expenditure for DL-WPCA was 354s, significantly less than the traditional 2436s.
• Memory overhead for DL-WPCA was 6.35s, compared to 186.25s for the traditional method.

Conclusions:

• Deep Learning-based WPCA is faster and more memory-efficient than traditional algorithms.
• The proposed methods enhance the accuracy and learning rate of web page classification.
• DL-based WPCA offers a more efficient solution for processing web page data.