Automatic Lossless-Summarization of News Articles with Abstract Meaning Representation

Mishra, Ritwik; Gayen, Tirthankar

doi:10.1016/j.procs.2018.08.164

Cited by 14 publications

(5 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some approaches form a pipeline by using the AMR representation as an intermediate step, where an input text is transformed into an AMR graph which is used for estimating a summary graph that, in turn, generates the final summary in textual form (Liu et al, 2018;Dohare et al, 2017Dohare et al, , 2018Mishra and Gayen, 2018). More specifically, in (Liu et al, 2018), the authors propose a framework for creating an AMR graph of a document by combining the sub-graphs of each sentence and transforming them into a single graph.…”

Section: Related Workmentioning

confidence: 99%

“…Then, a summary graph is created by finding the most important relations between nodes, capturing the surrounding information of each path in the graph. Furthermore, in (Mishra and Gayen, 2018), a co-reference resolution is performed before retrieving the AMR graphs of the sentences. Also, an algorithm for merging the AMR graphs is proposed to form the summary graph.…”

Section: Related Workmentioning

confidence: 99%

“…Also, an algorithm for merging the AMR graphs is proposed to form the summary graph. To generate the summary, the last three approaches described above, (Dohare et al, 2017), (Dohare et al, 2018) and (Mishra and Gayen, 2018), use an available AMR-to-text generator (Flanigan et al, 2016;Konstas et al, 2017). All of the aforementioned approaches employ a series of transformations; from text to text graph, from text graph to summary graph, and from summary graph to summary, increasing the overall computational load.…”

Section: Related Workmentioning

confidence: 99%

“…Our framework is compared with other relevant approaches (Takase et al, 2016;Dohare et al, 2017;Mishra and Gayen, 2018) and in the next section, we report the performance of the most relevant systems. As relevant approaches, we assume those TS systems that are based on AMR representation and also are capable of coping with a dataset of a large size, such as the CNN/DailyMail or Gigaword.…”

Section: Competitive Approachesmentioning

confidence: 99%

See 3 more Smart Citations

Text summarization based on semantic graphs: An abstract meaning representation graph-to-text deep learning approach

Kouris

Alexandridis

Stafylopatis

2022

Preprint

View full text Add to dashboard Cite

Nowadays, due to the constantly growing amount of textual information, automatic text summarization constitutes an important research area in natural language processing. In this work, we present a novel framework that combines semantic graph representations along with deep learning predictions to generate abstractive summaries of single documents, in an effort to utilize a semantic representation of the unstructured textual content in a machine-readable, structured, and concise manner. The overall framework is based on a well defined methodology for performing semantic graph parsing, graph construction, graph transformations for machine learning models and deep learning predictions. The employed semantic graph representation focuses on using the model of abstract meaning representation. Several combinations of graph construction and graph transformation methods are investigated to specify the most efficient of them for the machine learning models. Additionally, a range of deep learning architectures is examined, including a sequence-to-sequence attentive network, reinforcement learning, transformer-based architectures, and pre-trained neural language models. In this direction, a semantic graph representation of an original text is extracted, and then the present framework formulates the problem as a graph-to-summary learning problem to predict a summary of an original text. To the best of our knowledge, this formulation of graph-to-summary predictions in abstractive text summarization, without other intermediate steps in the machine learning phase, has not been presented in the relevant literature. Another important contribution is an extension of a metric for assessing the factual consistency of the generated summaries in an effort to provide a qualitative evaluation. To assess the framework, an extensive experimental procedure is presented that uses popular datasets to evaluate key aspects of the proposed approach. The obtained results exhibit promising performance, validating the robustness of the proposed framework.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

Section: Competitive Approachesmentioning

confidence: 99%

See 2 more Smart Citations

Text summarization based on semantic graphs: An abstract meaning representation graph-to-text deep learning approach

Kouris

Alexandridis

Stafylopatis

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…AMR cannot depict coreferences which cross sentence limits and, likewise, it cannot discern amongst hypothetical or real events. The available mention-pair systems whichever identify coreferences [3] do not analyze entity-level information when establishing if a couple of mentions are referent or not. Therefore, beginning from this ascertainment it was suggest an entity-centric coreference resolution system, which composes coreference chains step by step.…”

Section: Introductionmentioning

confidence: 99%

Automatic Comprehension of Geometry Problems using AMR Parser (S)

Iordan¹

2021

Proceedings of the 33rd International Conference on Software Engineering and Knowledge Engineering

View full text Add to dashboard Cite

Automatic comprehension of geometry problems described in natural language is a crucial and challenging stage of numerous automatic geometry problem solvers. These systems should comprehend the existing information in natural language geometry problems with the purpose to extract the geometric relationships between elements and to accomplish automatic solutions using intelligent methods. Abstract Meaning Representation is a popular framework for annotating whole sentence meaning. This paper proposes the addition of a new feature to an existing transition-based AMR parser that constructs AMR graphs from statement of geometry problems described in English language. The new feature consists in explicit embedding of the coreference detection into the parser. Integrating the automatic comprehension method with different geometry systems will greatly enhance the efficiency and intelligence in automatic solving. This approach shows improvements over the best previously published systems for extract information from statement of geometry problems described in English language.

show abstract

Stopwords Aware Emotion-Based Sentiment Analysis of News Articles

Yadav

Gayen

2023

EAI/Springer Innovations in Communication and Computing

View full text Add to dashboard Cite

Automatic Lossless-Summarization of News Articles with Abstract Meaning Representation

Cited by 14 publications

References 19 publications

Text summarization based on semantic graphs: An abstract meaning representation graph-to-text deep learning approach

Text summarization based on semantic graphs: An abstract meaning representation graph-to-text deep learning approach

Automatic Comprehension of Geometry Problems using AMR Parser (S)

Stopwords Aware Emotion-Based Sentiment Analysis of News Articles

Contact Info

Product

Resources

About