The continuing increasing of data amount make the traditional file management systems get into trouble in retrieval massive data information. Current mainstream file systems are focused on data access performance, reliability, but little attention costed to the correlation between data feature and data. According to the current file management system's existing disadvantages in file organization and retrieval, a fast query file management mode of coal quality informatization is Put forward under the coal enterprise's demand for file management. Analyzed the kinds of coal quality file in details, put forward a file organization way by integrating node with information and multi-dimensional document classification, adopt a topic relevance analysis algorithm and semantic similarity calculation to realize fast text retrieval. Finlly, through the design of protyping system and performance testing, the results show that the feasibility and efficiency of the theory is realized.