The invention discloses a method for news aggregation and intelligent entity association, which adopts crawler to grasp web pages by polling new news on websites interested by users, and utilizes support vector machine to carry out 0-1 classification to extract the news text, after the natural language processing of news text, the Wikidata knowledge map is searched for corresponding entities for characters and geographical names in the text, entity types are determined by superordinate words, news hexadecimals (title, time, URL, body, character entity, geographic entity) are stored in the local document database, the relevant news are listed when the user searches for the relevant entity, and the news location associated with Wikida is displayed by map, and the character file associated with Wikida is displayed by cards. The technical proposal of the invention provides an enhanced news reading mode with background knowledge pushed by the related knowledge, and improves the reading experience of the user.