Current computer systems have not yet worked in exact accordance with human cognition, perception, and knowledge organization. The gap between computer language and the human cognitive network becomes particularly apparent when computer systems are applied to literary research and education. In view of such subtle incompatibility, this study attempts to enhance a more cooperative application of computer systems to the literary field by establishing a semantic lexical database on the bases of textual contexts, linguistic characteristics, and semantics of Chinese poetry. In reference to conventional classifications, this semantic lexical database consists of four layers of semantic properties. The top layer is divided into 6 categories, including people, event, time, place, object, and others. The second layer is divided into 50 categories. The third layer has 368 subclasses. The fourth layer is composed of synonyms of individual subclasses. The design of this semantic lexical database improves the functionality of computer search engine as its operation shifts from information retrieval according to forms of characters to retrieval in terms of meanings of vocabulary items. This way of operation produces search results in a more precise manner that may meet the user’s demands for resources of specific themes. The enhancement of computer systems enables the search engine to function according to thinking modes of human brain and as a result constructs a structure similar to the human knowledge system. This advance may facilitate a more congruent application of computer systems to research and education in the field of literature.