An improved sentence polishing model used in automatic extraction

Language modeling plays a critical role for automatic extraction. Typically, the statistical model of automatic extraction suffers from the lack of the subject semantic consistency between sentences and the redundancy of information. In this study, we first introduce our work on automatic extraction, and then analyze the disadvantages of different extracting models. We then present a advanced mathematical model to overcome these lacks based on computational linguistics. As shown by experiments, the proposed modeling and methods can significantly reduce the redundancy of information and increase the subject semantic consistency between sentences of automatic abstraction with moderate computational cost.