Single Document Summarization

The availability of online information shows a need of efficient text summarization system. The text summarization system follows extractive and abstractive methods. In extractive summarization, the important sentences are selected from the original text on the basis of sentence ranking methods. The Abstractive summarization system understands the main concept of texts and predicts the overall idea about the topic. This paper mainly concentrated the survey of existing extractive text summarization models. Numerous algorithms are studied and their evaluations are explained. The main purpose is to observe the peculiarities of existing extractive summarization models and to find a good approach that helps to build a new text summarization system.

Focused Multi-Document Summarization (MDS) is concerned with summarizing documents in a collection with a concentration toward a particular external request (ie query, question, topic, etc.), or focus. Although the current state-of-the-art provides somewhat ...

In this paper a methodology to mine the concepts from documents and use these concepts to generate an objective summary of the claims section of the patent documents is proposed. Conceptual Graph (CG) formalism as proposed by Sowa (Sowa 1984) is used in this work for representing the concepts and their relationships. Automatic identification of concepts and conceptual relations from text documents is a challenging task. In this work the focus is on the analysis of the patent documents, mainly on the claim's section (Claim) of the documents. There are several complexities in the writing style of these documents as they are technical as well as legal. It is observed that the general in-depth parsers available in the open domain fail to parse the 'claims section' sentences in patent documents. The failure of in-depth parsers has motivated us, to develop methodology to extract CGs using other resources. Thus in the present work shallow parsing, NER and machine learning technique for extracting concepts and conceptual relationships from sentences in the claim section of patent documents is used. Thus, this paper discusses i) Generation of CG, a semantic network and ii) Generation of abstractive summary of the claims section of the patent. The aim is to generate a summary which is 30% of the whole claim section. Here we use Restricted Boltzmann Machines (RBMs), a deep learning technique for automatically extracting CGs. We have tested our methodology using a corpus of 5000 patent documents from electronics domain. The results obtained are encouraging and is comparable with the state of the art systems.

Current multi-document summarization systems can successfully extract summary sentences, however with
many limitations including: low coverage, inaccurate extraction to important sentences, redundancy and poor coherence among the selected sentences. The present study introduces a new concept of centroid approach and reports new techniques for extracting summary sentences for multi-document. In both
techniques keyphrases are used to weigh sentences and documents. The first summarization technique (Sen-Rich) prefers maximum richness sentences. While the second (Doc-Rich), prefers sentences from centroid document. To demonstrate the new summarization system application to extract summaries of Arabic documents we performed two experiments. First, we applied Rouge measure to compare the new techniques among systems presented at TAC2011. The results show that Sen-Rich outperformed all systems in ROUGE-S. Second, the system was applied to summarize multi-topic documents. Using human evaluators, the results show that Doc-Rich is the superior, where summary sentences characterized by extra coverage and more cohesion.

ABSTRACT Due to the exponential growth of textual information available on the Web, end users need to be able to access information in summary form – and without losing the most important information in the document when generating the summaries. Automatic generation of extractive summaries from a single document has traditionally been given the task of extracting the most relevant sentences from the original document. The methods employed generally allocate a score to each sentence in the document, taking into account certain features. The most relevant sentences are then selected, according to the score obtained for each sentence. These features include the position of the sentence in the document, its similarity to the title, the sentence length, and the frequency of the terms in the sentence. However, it has still not been possible to achieve a quality of summary that matches that performed by humans and therefore methods continue to be brought forward that aim to improve on the results. This paper addresses the generation of extractive summaries from a single document as a binary optimization problem where the quality (fitness) of the solutions is based on the weighting of individual statistical features of each sentence – such as position, sentence length and the relationship of the summary to the title, combined with group features of similarity between candidate sentences in the summary and the original document, and among the candidate sentences of the summary. This paper proposes a method of extractive single-document summarization based on genetic operators and guided local search, called MA-SingleDocSum. A memetic algorithm is used to integrate the own-population-based search of evolutionary algorithms with a guided local search strategy. The proposed method was compared with the state of the art methods UnifiedRank, DE, FEOM, NetSum, CRF, QCS, SVM, and Manifold Ranking, using ROUGE measures on the datasets DUC2001 and DUC2002. The results showed that MA-SingleDocSum outperforms the state of the art methods.

For effective multi-document summarization, it is important to reduce redundant information in the summaries and extract sentences, which are common to given documents. This paper presents a document summarization model which extracts key sentences from given documents while reducing redundant information in the summaries. An innovative aspect of our model lies in its ability to remove redundancy while selecting representative sentences. The model is represented as a discrete optimization problem. To solve the discrete optimization problem in this study an adaptive DE algorithm is created. We implemented our model on multi-document summarization task. Experiments have shown that the proposed model is to be preferred over summarization systems. We also showed that the resulting summarization system based on the proposed optimization approach is competitive on the DUC2002 and DUC2004 datasets.

For effective multi-document summarization, it is important to reduce redundant information in the summaries and extract sentences, which are common to given documents. This paper presents a document summarization model which extracts key sentences from given documents while reducing redundant information in the summaries. An innovative aspect of our model lies in its ability to remove redundancy while selecting representative sentences. The model is represented as a discrete optimization problem. To solve the discrete optimization problem in this study an adaptive DE algorithm is created. We implemented our model on multi-document summarization task. Experiments have shown that the proposed model is to be preferred over summarization systems. We also showed that the resulting summarization system based on the proposed optimization approach is competitive on the DUC2002 and DUC2004 datasets.

Automatic document summarization is always attractive to computer science researchers. A novel approach is proposed to address this topic and mainly focuses on the summarization of plain documents. Conventional summarization methods do not fully use the inter-sentence relevance that is not preserved during the processing. In contrast, to tackle the problem and incorporate the latent relations among sentences, our approach constructs relevance structures at sentence-level for plain documents and each sentence is scored with a significance value. Accordingly, important sentences “present” themselves automatically, and the summary paragraph is then generated by selecting top-k scored sentences. Convergence of the algorithm is proved, and experiment, which is conducted on two data sets (DUC 2006 and DUC 2007), shows that the proposed model gives convincing results.

Due to the exponential growth of textual information available on the Web, end users need to be able to access information in summary formand without losing the most important information in the document when generating the summaries. Automatic generation of extractive summaries from a single document has traditionally been given the task of extracting the most relevant sentences from the original document. The methods employed generally allocate a score to each sentence in the document, taking into account certain features. The most relevant sentences are then selected, according to the score obtained for each sentence. These features include the position of the sentence in the document, its similarity to the title, the sentence length, and the frequency of the terms in the sentence. However, it has still not been possible to achieve a quality of summary that matches that performed by humans and therefore methods continue to be brought forward that aim to improve on the results. This paper addresses the generation of extractive summaries from a single document as a binary optimization problem where the quality (fitness) of the solutions is based on the weighting of individual statistical features of each sentence -such as position, sentence length and the relationship of the summary to the title, combined with group features of similarity between candidate sentences in the summary and the original document, and among the candidate sentences of the summary. This paper proposes a method of extractive single-document summarization based on genetic operators and guided local search, called MA-SingleDocSum. A memetic algorithm is used to integrate the own-population-based search of evolutionary algorithms with a guided local search strategy. The proposed method was compared with the state of the art methods UnifiedRank, DE, FEOM, NetSum, CRF, QCS, SVM, and Manifold Ranking, using ROUGE measures on the datasets DUC2001 and DUC2002. The results showed that MA-SingleDocSum outperforms the state of the art methods.

Due to the exponential growth of textual information available on the Web, end users need to be able to access information in summary form-and without losing the most important information in the document when generating the summaries. Automatic generation of extractive summaries from a single document has traditionally been given the task of extracting the most relevant sentences from the original document. The methods employed generally allocate a score to each sentence in the document, taking into account certain features. The most relevant sentences are then selected, according to the score obtained for each sentence. These features include the position of the sentence in the document, its similarity to the title, the sentence length, and the frequency of the terms in the sentence. However, it has still not been possible to achieve a quality of summary that matches that performed by humans and therefore methods continue to be brought forward that aim to improve on the results. This paper addresses the generation of extractive summaries from a single document as a binary optimization problem where the quality (fitness) of the solutions is based on the weighting of individual statistical features of each sentence-such as position, sentence length and the relationship of the summary to the title, combined with group features of similarity between candidate sentences in the summary and the original document, and among the candidate sentences of the summary. This paper proposes a method of extractive single-document summarization based on genetic operators and guided local search, called MA-Sin-gleDocSum. A memetic algorithm is used to integrate the own-population-based search of evolutionary algorithms with a guided local search strategy. The proposed method was compared with the state of the art methods UnifiedRank, DE, FEOM, NetSum, CRF, QCS, SVM, and Manifold Ranking, using ROUGE measures on the datasets DUC2001 and DUC2002. The results showed that MA-SingleDocSum outperforms the state of the art methods.

Due to the exponential growth of textual information available on the Web, end users need to be able to access information in summary form-and without losing the most important information in the document when generating the summaries. Automatic generation of extractive summaries from a single document has traditionally been given the task of extracting the most relevant sentences from the original document. The methods employed generally allocate a score to each sentence in the document, taking into account certain features. The most relevant sentences are then selected, according to the score obtained for each sentence. These features include the position of the sentence in the document, its similarity to the title, the sentence length, and the frequency of the terms in the sentence. However, it has still not been possible to achieve a quality of summary that matches that performed by humans and therefore methods continue to be brought forward that aim to improve on the results. This paper addresses the generation of extractive summaries from a single document as a binary optimization problem where the quality (fitness) of the solutions is based on the weighting of individual statistical features of each sentence-such as position, sentence length and the relationship of the summary to the title, combined with group features of similarity between candidate sentences in the summary and the original document, and among the candidate sentences of the summary. This paper proposes a method of extractive single-document summarization based on genetic operators and guided local search, called MA-Sin-gleDocSum. A memetic algorithm is used to integrate the own-population-based search of evolutionary algorithms with a guided local search strategy. The proposed method was compared with the state of the art methods UnifiedRank, DE, FEOM, NetSum, CRF, QCS, SVM, and Manifold Ranking, using ROUGE measures on the datasets DUC2001 and DUC2002. The results showed that MA-SingleDocSum outperforms the state of the art methods.

Focused Multi-Document Summarization (MDS) is concerned with summarizing documents in a collection with a concentration toward a particular external request (i.e. query, question, topic, etc.), or focus. Although the current state-of-the-art provides somewhat decent performance for DUC/TAC-like evaluations (i.e. government and news concerns), other considerations need to be explored. This paper not only briefly explores the state-of-the-art in automatic systems techniques, but also a comparison with human summarization activity.

In this paper a methodology to mine the concepts from documents and use these concepts to generate an objective summary of the claims section of the patent documents is proposed. Conceptual Graph (CG) formalism as proposed by Sowa (Sowa 1984) is used in this work for representing the concepts and their relationships. Automatic identification of concepts and conceptual relations from text documents is a challenging task. In this work the focus is on the analysis of the patent documents, mainly on the claim's section (Claim) of the documents. There are several complexities in the writing style of these documents as they are technical as well as legal. It is observed that the general in-depth parsers available in the open domain fail to parse the 'claims section' sentences in patent documents. The failure of in-depth parsers has motivated us, to develop methodology to extract CGs using other resources. Thus in the present work shallow parsing, NER and machine learning technique for extracting concepts and conceptual relationships from sentences in the claim section of patent documents is used. Thus, this paper discusses i) Generation of CG, a semantic network and ii) Generation of abstractive summary of the claims section of the patent. The aim is to generate a summary which is 30% of the whole claim section. Here we use Restricted Boltzmann Machines (RBMs), a deep learning technique for automatically extracting CGs. We have tested our methodology using a corpus of 5000 patent documents from electronics domain. The results obtained are encouraging and is comparable with the state of the art systems.