The invention discloses a patent personalized retrieval analysis system based on big data. The system comprises a retrieval module, a personalization module and an analysis module. The personalized module serves the retrieval module and the analysis module; the personalized module comprises a user model unit, a data acquisition unit, a learning unit, a reminding unit and a storage unit, the retrieval module comprises a field distinguishing unit, a keyword searching unit and a theme unit, and the analysis module comprises an automatic distribution unit, a file comparison unit, a duplicate checking unit and an innovation level analysis unit. According to the present invention, patents are evaluated and analyzed from many aspects, for example including the manual judgment and the systematic duplicate checking, so that patent analysis has subjectivity and objectivity, and the patent analysis is more comprehensive. By carrying out the personalized analysis from the aspects, such as the userhabits, the historical behaviors, etc., the requirements of the user are better known, and the latest trend of patents in the field is known in time through the attention of the user and by pushing the related information.