A progressive sentence selection strategy for document summarization

Ouyang, You; Li, Wenjie; Zhang, Renxian; Li, Sujian; Lu, Qin

doi:10.1016/j.ipm.2012.05.002

Cited by 22 publications

(16 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The function of similarity which is used is uni-gram matching-based similarity based on the equation (1). Similarity between sentences is calculated based on corresponding words between the words s to-i and the words s to-j (…”

Section: Sentence Clustering Phase With Similarity Based Histogram CLmentioning

confidence: 99%

“…Good document summary is a summary of which is capable of covering (coverage) as much as possible the important concepts (saliency) that exist in the source document [1] and saliency are a major problem in the document summary, the strategy to select a sentence is very important because it should be able to choose the main phrases and avoid redundancy so as to include many of the concepts [2]. Some studies [2][3][4][5] have developed a method of selecting an important sentence to address the issue of coverage and saliency.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Improving Multi-Document Summary Method Based on Sentence Distribution

2016

View full text Add to dashboard Cite

show abstract

Section: Sentence Clustering Phase With Similarity Based Histogram CLmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Improving Multi-Document Summary Method Based on Sentence Distribution

2016

View full text Add to dashboard Cite

show abstract

“…Abstractive techniques generate summaries based on synthesized information units that may not be necessarily part of the original text. Most of the existing techniques are based on extractive summarization [9], [10], [11], [12] and are therefore considered in this paper. In other words, we do not deal with abstractive summarization.…”

Section: Introductionmentioning

confidence: 99%

“…An important concern in extractive summarization is to select important sentences or paragraphs [1], [5], [11], [3]. Several approaches have been proposed for this purpose such as sentence selection based on lexical chain distribution [13], [1], [3], [14], [15], similarity with the topic [9], [10], [11], [12] and classifying sentences as positive or negative with respect to the summary [16], [17], [18], [19], [11].…”

Section: Introductionmentioning

confidence: 99%

Text summarization using rough sets

Azam

Ahmad

2016

2016 International Conference on Computing, Electronic and Electrical Engineering (ICE Cube)

View full text Add to dashboard Cite

Text Summarization aims to generate concise and compressed form of original documents. The techniques used for text summarization may be categorized as extractive summarization and abstractive summarization. We consider extractive techniques which are based on selection of important sentences within a document. A major issue in extractive summarization is how to select important sentences, i.e., what criteria should be defined for selection of sentences which are eventually part of the summary. We examine this issue using rough sets notion of reducts. A reduct is an attribute subset which essentially contains the same information as the original attribute set. In particular, we defined and examined three types of matrices based on an information table, namely, discernibility matrix, indiscernibility matrix and equal to one matrix. Each of these matrices represents a certain type of relationship between the objects of an information table. Three types of reducts are determined based on these matrices. The reducts are used to select sentences and consequently generate text summaries. Experimental results and comparisons with existing approaches advocates for the use of the proposed approach in generating text summaries.

show abstract

“…Salience and coverage are two important issues for evaluating the quality of extractive summaries [1]. A salient summary should contain the sentences that are most relevant to the main topics of original document.…”

Section: Introductionmentioning

confidence: 99%

Text Summarization Based on Sentence Selection with Semantic Representation

Zhang

Wang

et al. 2014

2014 IEEE 26th International Conference on Tools With Artificial Intelligence

View full text Add to dashboard Cite

Text summarization is of great importance to solve information overload. Salience and coverage are two most important issues for summaries. Most existing models extract summaries by selecting the top sentences with highest scores without using the relationships between sentences, and usually represent the sentences simply basing on lexical or statistical features. As a result, those models can not achieve salience or coverage very well. In this paper, we propose a novel summarization model called Sentence Selection with Semantic Representation (SSSR). SSSR ensures both salience and coverage by learning semantic representations for sentences and applying a well-designed selection strategy to select summary sentences. The selection strategy used in SSSR is to select sentences that can reconstruct the original document with least distortion by means of linear combination. Besides, we improve our selection strategy by reducing redundant information. Then we learn two semantic representations for sentences: (1) weighted mean of word embeddings; (2) deep coding. Both of them are semantic and compact, and can capture similarities between sentences. Extensive experiments on datasets DUC2006 and DUC2007 validate our model.

show abstract

A progressive sentence selection strategy for document summarization

Cited by 22 publications

References 15 publications

Improving Multi-Document Summary Method Based on Sentence Distribution

Improving Multi-Document Summary Method Based on Sentence Distribution

Text summarization using rough sets

Text Summarization Based on Sentence Selection with Semantic Representation

Contact Info

Product

Resources

About