基于连续LexRank的多文本自动摘要优化算法研究

RESEARCH ON OPTIMISATION ALGORITHM OF MULTIPLE TEXT AUTO-SUMMARISATION BASED ON CONTINUOUS LEXRANK

以挖掘性的自动摘要(TS)为研究对象,依赖于核心语句这一概念,在考虑现存相关研究成果的基础上,设计一种基于特征向量中心概念及连续LexRank、以图形表示的多文本自动摘要优化模型及算法。在此模型中,创建了一个基于内语句余弦相似度连接矩阵以实现语句的图形表示形式对应的邻接矩阵。为了验证算法的可行性与效率,设计了相关实验方案,并通过与现存算法执行效果进行实时比对。实验结果表明,提出的带阈值及基于连续LexRank的算法具有较高的效率。

We take the text automatic summarisation （TAS） with mining property as the study object, rely on the concept of salient sentence, based on taking into account the existing correlated research outcomes, we design a multiple text automatic summarisation optimisation model and algorithm. The model is based on the concept of eigenvector centrality and continuous LexRank, and is represented in graphics. In this model, a connectivity matrix based on intra-sentence cosine similarity is constructed to realise the adjacency matrix corresponding to the graph representation of sentences. In order to verify the feasibility and efficiency of the algorithm, we design the correlated experimental scheme, and make. real-time comparison with the execution effect of current algorithm. Experimental result demonstrates that the algorithm proposed in the paper with threshold and based on continuous LexRank has higher efficiency.