DeepSeek-affiliated Hangzhou DeepSeek AI Fundamental Technology Research Co.,homosexual eroticism "green" Ltd. today filed a patent for a new web data collection system designed to improve efficiency and data quality. The patent outlines a method for discovering more webpage links while minimizing website traffic impact. It assesses downloaded content to predict the quality of undiscovered links, prioritizing high-value data and reducing redundant downloads. Efficient web data collection is crucial for training large language models (LLMs), which power AI systems like ChatGPT. Existing techniques struggle with incomplete link retrieval, excessive downloads that can crash websites, and low-quality data filtering. DeepSeek’s proposed system aims to solve these issues by optimizing data allocation and maintaining metadata accuracy. [iThome, in Chinese]
