BestBlogs.dev
Gather top programming, artificial intelligence, products, and technology articles, large language model summary score assists reading, and explore the future of programming and technology.
introduce
bestblogs.dev provides you with unique value-oriented content in programming, artificial intelligence, product design, business technology and personal growth, bringing together top technology companies and communities. We use advanced language models to summarize, rate and translate these articles for you, saving you time reading. We understand the pain points of data screening and are committed to presenting you with selected content. Subscribe now and explore the endless possibilities of future technology!

subscription
Website subscription address: https://www.bestblogs.dev/#subscribe
The latest selected pushes of this week will be pushed every Friday, so that you can get to know the latest artificial intelligence development technologies and trends at the first time.

RSS Source
The website articles are sourced from all the following RSS feeds (200):
https://github.com/ginobefun/BestBlogs/blob/main/BestBlogsRSS.opml
Among them, the WeChat public account to RSS is built using the wewe-rss project. The current supported WeChat public account RSS subscription sources (200) are sorted out as follows:
https://github.com/ginobefun/BestBlogs/blob/main/WeWeRSS.opml
You can browse all RSS feed information on the web page https://www.bestblogs.dev/sources, including the number of articles, selected articles, and reads in the last 3 months of the feed.

You can use RSS readers to import and read. If you have high-quality RSS feeds that need to be added, you can add Issue.
Implementation principle
BestBlogs.dev article intelligent analysis practice based on Dify Workflow

- Article crawling process: Based on the RSS protocol, crawl article information from all subscribers, including title, link, publishing time, etc., and crawl article content through links and headless browsers. The text is extracted through the article body selector, etc. defined on the source, and the HTML, pictures, etc. of the text are processed, and placed in the list of pending articles.
- The initial review process of the article: Through the characteristics of language, article content and other characteristics, the article is initially scored, low-quality articles are eliminated, and subsequent steps are reduced. Use the Dify Workflow project to conduct initial review of articles. For detailed instructions, please refer to the BestBlogs article initial review process.
- Article analysis process: The article is summarized, classified and scored through a large language model (such as GPT-4o), and a sentence summary, article summary, main opinions, article golden sentences, fields, label lists and scores are generated, so that readers can quickly filter and filter, understand the main content of the full text, and judge whether to continue reading. Use the Dify Workflow project to perform article analysis, including segmented analysis - summary analysis - domain division and label generation - article rating - inspection and reflection - optimization and improvement, etc. For detailed instructions, please refer to BestBlogs article analysis process
- Translation process of article analysis results: Translate the article analysis results through a large language model (such as GPT-4o). Currently, the website supports two languages, Chinese and English, and generates abstracts, main views, article golden sentences, label lists, etc. of the target language based on the language of the original text. Use the Dify Workflow project to translate article analysis results, including identifying professional terms - initial translation - checking translation - free translation and other links. For detailed instructions, please refer to the BestBlogs article analysis results translation process.
Acknowledge