In the internet stock news crawl work, should be based on thea target data model.; Aaccess toing certain patterns of iInternet stock news webpage data, and transforming it, mapping it into the desired target data pattern. According to the stock news website sSource cCode dDesign rRules', rules for web crawler collection, stock news crawl only needs to intercept: the company ID, news content, author, date of publication, and last updated time - in five parts, as shown in screenshot 1. This study used a self-written crawler program designed in pPython programming language, given in the initial website.