Formulation of document summarization as a 0–1 nonlinear programming problem

► We model text summarization as a nonlinear 0–1 programming problem. ► This model balances content coverage and diversity in the summary. ► We utilize the DPSO-EDA algorithm to solve the optimization problem. ► Experiments show that our model produces very competitive results. We proposed a novel t...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Computers & industrial engineering Ročník 64; číslo 1; s. 94 - 102
Hlavní autoři: Alguliev, Rasim M., Aliguliyev, Ramiz M., Isazade, Nijat R.
Médium: Journal Article
Jazyk:angličtina
Vydáno: New York Elsevier Ltd 01.01.2013
Pergamon Press Inc
Témata:
Balancing coverage and diversity
Discrete particle swarm optimization
Documentation
Estimation of distribution
Generic summarization
Linear programming
Mathematical models
Mathematical problems
Nonlinear programming
Optimization model
Studies
Syntax
Weighted combination of arithmetic and geometric means
Discrete particle swarm optimization
Estimation of distribution
Weighted combination of arithmetic and geometric means
Generic summarization
Optimization model
Balancing coverage and diversity
ISSN:0360-8352, 1879-0550
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:► We model text summarization as a nonlinear 0–1 programming problem. ► This model balances content coverage and diversity in the summary. ► We utilize the DPSO-EDA algorithm to solve the optimization problem. ► Experiments show that our model produces very competitive results. We proposed a novel text summarization model based on 0–1 non-linear programming problem. This proposed model covers the main content of the given document(s) through sentence assignment. We implemented our model on multi-document summarization task. When comparing our method to several existing summarization methods on an open DUC2001 and DUC2002 datasets, we found that the proposed method could improve the summarization results significantly. The methods were evaluated using ROUGE-1, ROUGE-2 and ROUGE-W metrics.
Bibliografie:SourceType-Scholarly Journals-1
ObjectType-Feature-1
content type line 14
ISSN:0360-8352
1879-0550
DOI:10.1016/j.cie.2012.09.005