NLP-文本摘要-2004：LexRank（无监督图模型）【属于机器学习范畴；由PageRank网页重要性分数衍生而来；用于文本“关键词”、 “关键句” 提取、抽取式摘要】

àì夳堔傛蜴生んèń 2023-10-01 11:12 239阅读 0赞

[《原始论文：LexRank: Graph-based Lexical Centrality as Salience in Text Summarization》][LexRank_ Graph-based Lexical Centrality as Salience in Text Summarization]

![在这里插入图片描述][watermark_type_d3F5LXplbmhlaQ_shadow_50_text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO_size_20_color_FFFFFF_t_70_g_se_x_16]  
![在这里插入图片描述][watermark_type_d3F5LXplbmhlaQ_shadow_50_text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO_size_20_color_FFFFFF_t_70_g_se_x_16 1]  
句子Si与Sj之间的相似度使用余弦相似度作为相似度计算方式，

如图所示：

![在这里插入图片描述][watermark_type_d3F5LXplbmhlaQ_shadow_50_text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO_size_20_color_FFFFFF_t_70_g_se_x_16 2]

LexRank是一种无监督的基于图形的算法，它使用IDF修改的余弦作为两个句子之间的相似性度量。该相似度用作两个句子之间的图形边缘的权重。LexRank还采用了智能后处理步骤，确保为摘要选择的顶级句子彼此不太相似。

--------------------

参考资料：  
[Lexrank学习][Lexrank]

[LexRank_ Graph-based Lexical Centrality as Salience in Text Summarization]: https://arxiv.org/abs/1109.2128
[watermark_type_d3F5LXplbmhlaQ_shadow_50_text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO_size_20_color_FFFFFF_t_70_g_se_x_16]: https://img-blog.csdnimg.cn/83ed79f0f88d462a80b6ade95b71c1a0.png?x-oss-process=image/watermark,type_d3F5LXplbmhlaQ,shadow_50,text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO,size_20,color_FFFFFF,t_70,g_se,x_16
[watermark_type_d3F5LXplbmhlaQ_shadow_50_text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO_size_20_color_FFFFFF_t_70_g_se_x_16 1]: https://img-blog.csdnimg.cn/4ebdda8c2bed4bada636379c773cba7c.png?x-oss-process=image/watermark,type_d3F5LXplbmhlaQ,shadow_50,text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO,size_20,color_FFFFFF,t_70,g_se,x_16
[watermark_type_d3F5LXplbmhlaQ_shadow_50_text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO_size_20_color_FFFFFF_t_70_g_se_x_16 2]: https://img-blog.csdnimg.cn/6ab6d62a21934be6985471b1dbf517bf.png?x-oss-process=image/watermark,type_d3F5LXplbmhlaQ,shadow_50,text_Q1NETiBA5b-N6ICF44Gu5Lmx5aSq6YOO,size_20,color_FFFFFF,t_70,g_se,x_16
[Lexrank]: https://blog.csdn.net/Silience_Probe/article/details/80699986