Title :
An improvement on approximate dynamic programming for multi-document summarization
Author :
Yihui Luo ; Shuchu Xiong
Author_Institution :
Coll. of Comput. & Inf., Hunan Univ. of Commerce ChangshaCollege of Computer and Informationy, Changsha, China
Abstract :
A new improved approximate dynamic programming for multi-document summarization is presented. Our proposed algorithm improves the state-of-art approximate dynamic programming algorithm for multi-document summarization in [1]. The improvement of our method is attributed to the adding search in the backward direction at each sequential step of the dynamic programming procedure. The experimental results for multi-document summarization tasks on DUC2002 and DUC2004 data sets validate the effectiveness of our proposed method.
Keywords :
document handling; dynamic programming; DUC2002 data set; DUC2004 data set; dynamic programming algorithm; multidocument summarization; optimization; Algorithm design and analysis; Approximation algorithms; Data models; Dynamic programming; Heuristic algorithms; Linear programming; Optimization; backward search; dynamic programming algorithm; multi-document summrization; optimization problem;
Conference_Titel :
Security, Pattern Analysis, and Cybernetics (SPAC), 2014 International Conference on
Conference_Location :
Wuhan
Print_ISBN :
978-1-4799-5352-3
DOI :
10.1109/SPAC.2014.6982716
