The rapid progress of digitalization of ancient literatures in China has promoted the excavation and dissemination of traditional Chinese culture dynamically. In the digital era, the collation of ancient literatures should be upgraded from static retrieval to dynamic characterization, and the use of new technologies to improve the service efficiency of ancient literatures is of great significance to the transformation and development of ancient literatures' classification. Based on the theory of "World 3", this research conduct a analytical procedure of word segmentation and word frequency on Shi Ji (Historical Records) via Jiayan and Hidden Markov Model and then makes a visual analysis of the full text distribution of some high-frequency words, taking into account the collocation of surrounding related words, so as to provide readers with more comprehensive linguistic information. The procedures of Natural Language Processing conduct by this research on ancient literatures is expected to provide a possible reference for the knowledge organization of ancient books.