篇名 | A Maximum Entropy model for Automatic Summarization |
---|---|
卷期 | 27:2 |
作者 | Wei-Jiang Li 、 Zhen-Zhen Wang 、 Dong-Jun Li 、 Zheng-Tao Yu 、 Tie-Jun Zhao |
頁次 | 050-058 |
關鍵字 | Automatic Summarization 、 Maximum Entropy 、 Feature Extraction 、 EI 、 MEDLINE 、 Scopus |
出刊日期 | 201607 |
Nowadays, the number of electronic information has grown largely. Research on summarization is particularly important. Automatic summarization can accelerate the speed of access to resources. This paper studied the automatic summarization existing technical methods and the basic principle of maximum entropy model. With the principle of maximum entropy and automatic summarization technical characteristics, the au-tomatic summarization method based on maximum entropy model is designed. Summary sentences multifac-eted characteristics, design the automatic summarization sentences characteristics of maximum entropy ex-traction rules, different digest the results of different combinations of features. The results of experiments and examples show that the new method has a good practical effect.