The Design And Implementation Of Ancient Poetry Knowledge Search System

As a literature resource platform, CADAL Chinese Chronicle Literature System is still in the primary stage. How to deepen the knowledge of the existing system and provide users with more convenient, more abundant, deeper and more interesting knowledge is a turning point of the system from the primary to the mature stage.In this paper, as the starting point of ancient poetry and the appreciation, we use text data mining, information retrieval and other related technologies to build an ancient poetry knowledge search system. The main work is as follows:1. Text analysis algorithms are proposed for the ancient poetry knowledge search, including:(1) uncontrolled keywords extraction algorithm based on word extraction, language model and clustering is used to achieve the keywords extraction of ancient poetry and support system theme search; (2) controllable keywords automatic labeling algorithm based on word distance calculation and rule optimization is used to realize the extraction of implicit knowledge from ancient poetry data and support system theme search; (3) keywords semantic generalization method based on synonyms table is used to support system theme search; (4) universal and semi-supervised character relation mining method based on Chinese Wikipedia data is used to support system relation search.2. The overall framework of the automatic generation system of theme poetry anthology is builded, the logical architecture and physical architecture of the system are designed and the system is implemented. In particular, we collect different types of data from various data source to complete data aggregation. Then we extract and analyze knowledge by text analysis, image analysis and other techniques to provide support for system services.