Extractive Summarization Model Research Articles

Hybrid model for extractive single document summarization: utilizing BERTopic and BERT model

Extractive text summarization has been a popular research area for many years. The goal of this task is to generate a compact and coherent summary of a given document, preserving the most important information. However, current extractive summarization methods still face several challenges such as semantic drift, repetition, redundancy, and lack of coherence. A novel approach is presented in this paper to improve the performance of an extractive summarization model based on bidirectional encoder representations from transformers (BERT) by incorporating topic modeling using the BERTopic model. Our method first utilizes BERTopic to identify the dominant topics in a document and then employs a BERT-based deep neural network to extract the most salient sentences related to those topics. Our experiments on the cable news network (CNN)/daily mail dataset demonstrate that our proposed method outperforms state-of-the-art BERT-based extractive summarization models in terms of recall-oriented understudy for gisting evaluation (ROUGE) scores, which resulted in an increase of 32.53% of ROUGE-1, 47.55% of ROUGE-2, and 16.63% of ROUGE-L when compared to baseline BERT-based extractive summarization models. This paper contributes to the field of extractive text summarization, highlights the potential of topic modeling in improving summarization results, and provides a new direction for future research.

A study of extractive summarization of long documents incorporating local topic and hierarchical information

In recent years, the transformer-based language models have achieved remarkable success in the field of extractive text summarization. However, there are still some limitations in this kind of research. First, the transformer language model usually regards the text as a linear sequence, ignoring the inherent hierarchical structure information of the text. Second, for long text data, traditional extractive models often focus on global topic information, which poses challenges in how they capturing and integrating local contextual information within topic segments. To address these issues, we propose a long text extractive summarization model that employs a local topic information extraction module and a text hierarchical extraction module to capture the local topic information and document's hierarchical structure information of the original text. Our approach enhances the ability to determine whether a sentence belongs to the summary. In this experiment, ROUGE score is used as the experimental evaluation index, and evaluates the model on three large public datasets. Through experimental validation, the model demonstrates superior performance in terms of ROUGE-1, ROUGE-2, and ROUGE-L scores compared to current mainstream summarization models, affirming the effectiveness of incorporating local topic information and document hierarchical structure into the model.

Pre-Training Clustering Models to Summarize Vietnamese Texts

Our investigation aims at pre-training clustering models to summarize Vietnamese texts. For this purpose, we create a large-scale dataset by collecting Vietnamese articles from newspaper websites and extracting the plain text to build the dataset, including 1,101,101 documents. We propose a new single-document extractive text summarization model based on clustering models. Our proposal clusters the documents with the hard clustering k-means algorithm and the soft clustering LDA (Latent Dirichlet Allocation) algorithm. Then, based on the pre-training clustering models, a summary model is used to select the salient sentence in the input text to construct the summary. The empirical results showed that our summary model achieved 51.22% ROUGE-1, 17.62% ROUGE-2 and 29.16% ROUGE-L on the testing set. Besides the traditional word representation such as BoW (Bag-of-Words), we also use the word meaning-based tools like FastText and BERT (Bidirectional Encoder Representations from Transformers) in our model. The additional benefit of our proposed extractive summary model is that the output summary is a long-text, readable document. Furthermore, the model’s architecture is straightforward, easy to understand and runs on cost-efficient resources like arm CPU and GPU too.

Unsupervised Extractive Summarization with Learnable Length Control Strategies

Unsupervised extractive summarization is an important technique in information extraction and retrieval. Compared with supervised method, it does not require high-quality human-labelled summaries for training and thus can be easily applied for documents with different types, domains or languages. Most of existing unsupervised methods including TextRank and PACSUM rely on graph-based ranking on sentence centrality. However, this scorer can not be directly applied in end-to-end training, and the positional-related prior assumption is often needed for achieving good summaries. In addition, less attention is paid to length-controllable extractor, where users can decide to summarize texts under particular length constraint. This paper introduces an unsupervised extractive summarization model based on a siamese network, for which we develop a trainable bidirectional prediction objective between the selected summary and the original document. Different from the centrality-based ranking methods, our extractive scorer can be trained in an end-to-end manner, with no other requirement of positional assumption. In addition, we introduce a differentiable length control module by approximating 0-1 knapsack solver for end-to-end length-controllable extracting. Experiments show that our unsupervised method largely outperforms the centrality-based baseline using a same sentence encoder. In terms of length control ability, via our trainable knapsack module, the performance consistently outperforms the strong baseline without utilizing end-to-end training. Human evaluation further evidences that our method performs the best among baselines in terms of relevance and consistency.

Interpretable extractive text summarization with meta-learning and BI-LSTM: A study of meta learning and explainability techniques

Text summarization is a widely-researched problem among scholars in the field of natural language processing. Multiple techniques have been proposed to help tackle this problem, yet some of these methodologies may still exhibit limitations such as the requirements for large training datasets, which might not always be possible, but more importantly, the lack of interpretability or transparency of the model. In this paper, we propose using meta-learning algorithm to train a deep learning model for extractive text summarization and then using various explanatory techniques such as SHAP (Shapley, 1953), linear regression (Lederer, 2022), decision trees (Fürnkranz, 2010), and input modification to gain insights into the model’s decision making process. To evaluate the effectiveness of our approach, we will compare it to other popular natural language processing models like BERT (Miller, 2019) or XLNET (Yang et al., 2020) using the ROUGE metrics (Lin, 2004). Overall, our proposed approach provides a promising solution to the limitations of existing methods and a framework for improving the explainability of deep learning models in other natural language processing tasks.

Retracted: N-GPETS: Neural Attention Graph-Based Pretrained Statistical Model for Extractive Text Summarization.

[This retracts the article DOI: 10.1155/2022/6241373.].

Extractive social media text summarization based on MFMMR-BertSum

The advancement of computer technology has led to an overwhelming amount of textual information, hindering the efficiency of knowledge intake. To address this issue, various text summarization techniques have been developed, including statistics, graph sorting, machine learning, and deep learning. However, the rich semantic features of text often interfere with the abstract effects and lack effective processing of redundant information. In this paper, we propose the Multi-Features Maximal Marginal Relevance BERT (MFMMR-BertSum) model for Extractive Summarization, which utilizes the pre-trained model BERT to tackle the text summarization task. The model incorporates a classification layer for extractive summarization. Additionally, the Maximal Marginal Relevance (MMR) component is utilized to remove information redundancy and optimize the summary results. The proposed method outperforms other sentence-level extractive summarization baseline methods on the CNN/DailyMail dataset, thus verifying its effectiveness.

An Extractive Text Summarization Model for Generating Extended Abstracts of Medical Papers in Turkish

The rapid growth of technology has led to an increase in the amount of data available in the digital environment. This situation makes it difficult for users to find the information they are looking for within this vast dataset, making it time-consuming. To alleviate this difficulty, automatic text summarization systems have been developed as a more efficient way to access relevant information in texts compared to traditional summarization techniques. This study aims to extract extended summaries of Turkish medical papers written about COVID-19. Although scientific papers already have abstracts, more comprehensive summaries are still needed. To the best of our knowledge, automatic summarization of academic studies related to COVID-19 in the Turkish language has not been done before. A dataset was created by collecting 84 Turkish papers from DergiPark. Extended summaries of 2455 and 1708 characters were obtained using widely used extractive methods such as Term Frequency and LexRank algorithms, respectively. The performance of the text summarization model was evaluated based on Recall, Precision, and F-score criteria, and the algorithms were shown to be effective for Turkish. The results of the study showed similar accuracy rates to previous studies in the literature.

MRC-Sum: An MRC framework for extractive summarization of academic articles in natural sciences and medicine

Extractive summarization for academic articles in natural sciences and medicine has attracted attention for a long time. However, most existing extractive summarization models often process academic articles with sentence classification models, which are hard to produce comprehensive summaries. To address this issue, we explore a new view to solve the extractive summarization of academic articles in natural sciences and medicine by taking it as a question-answering process. We propose a novel framework, MRC-Sum, where the extractive summarization for academic articles in natural sciences and medicine is cast as an MRC (Machine Reading Comprehension) task. To instantiate MRC-Sum, article-summary pairs in the summarization datasets are firstly reconstructed into (Question, Answer, Context) triples in the MRC task. Several questions are designed to cover the main aspects (e.g. Background, Method, Result, Conclusion) of the articles in natural sciences and medicine. A novel strategy is proposed to solve the problem of the non-existence of the ground truth answer spans. Then MRC-Sum is trained on the reconstructed datasets and large-scale pre-trained models. During the inference stage, four answer spans of the predefined questions are given by MRC-Sum and concatenated to form the final summary for each article. Experiments on three publicly available benchmarks, i.e., the Covid, PubMed, and arXiv datasets, demonstrate the effectiveness of MRC-Sum. Specifically, MRC-Sum outperforms advanced extractive summarization baselines on the Covid dataset and achieves competitive results on the PubMed and arXiv datasets. We also propose a novel metric, COMPREHS, to automatically evaluate the comprehensiveness of the system summaries for academic articles in natural sciences and medicine. Abundant experiments are conducted and verified the reliability of the proposed metric. And the results of the COMPREHS metric show that MRC-Sum is able to generate more comprehensive summaries than the baseline models.

A Multi-Granularity Heterogeneous Graph for Extractive Text Summarization

Extractive text summarization selects the most important sentences from a document, preserves their original meaning, and produces an objective and fact-based summary. It is faster and less computationally intensive than abstract summarization techniques. Learning cross-sentence relationships is crucial for extractive text summarization. However, most of the language models currently in use process text data sequentially, which makes it difficult to capture such inter-sentence relations, especially in long documents. This paper proposes an extractive summarization model based on the graph neural network (GNN) to address this problem. The model effectively represents cross-sentence relationships using a graph-structured document representation. In addition to sentence nodes, we introduce two nodes with different granularity in the graph structure, words and topics, which bring different levels of semantic information. The node representations are updated by the graph attention network (GAT). The final summary is obtained using the binary classification of the sentence nodes. Our text summarization method was demonstrated to be highly effective, as supported by the results of our experiments on the CNN/DM and NYT datasets. To be specific, our approach outperformed baseline models of the same type in terms of ROUGE scores on both datasets, indicating the potential of our proposed model for enhancing text summarization tasks.

Multiple View Summarization Framework for Social Media

Social Media provide voluminous posts about current topics and events. When a user desires to investigate a popular topic, it is not feasible as there are many posts. Besides, posts show different biases, viewpoints, perspectives, and emotions. Thus, providing summaries of large post sets with different viewpoints is necessary. We develop a multiple view summa-rization framework to generate different view-based summar-ies of Twitter posts. Users can apply different methods to generate summaries: 1) Entity-centered, 2) Social feature-based, 3) Event-based summarization, using all triple embed-dings and 4) Sentiment-based summarization to generate summaries of positive or negative views of tweets. These summarization methods are compared with BertSum, SBert, T5, and Bart-Large-CNN with a gold standard dataset. Our results, based on Rouge scores, were better than these pub-lished extractive and abstractive summarization models.

Auto-regressive extractive summarization with replacement

Auto-regressive extractive summarization approaches determine sentence extraction probability conditioning on previous decisions by maintaining a partial summary representation. Despite its popularity, the framework has two main drawbacks: 1) the partial summary representation is irresolutely denoted by a weighted summation of all the processed sentences without any filtering, resulting in a noisy representation and degrading the effectiveness of extracting subsequent sentences; 2) earlier sentences are biased towards a higher extraction probability due to the sequential nature of sequence tagging. To address these two problems, we propose the Auto-regressive Extractive Summarization with Replacement (AES-Rep), a novel auto-regressive extractive summarization model. In particular, the AES-Rep model consists of two main modules: the extraction decision module that determines whether a sentence should be extracted, and the replacement locater module that enables extracted deficient sentences to be replaced with latter sentences by comparing their expressiveness with respect to the main idea of the document. These modules update the partial summary with explicit actions using elaborated multidimensional guidance. We conduct extensive experiments on the benchmark CNN and DailyMail datasets. Experimental results show that AES-Rep can achieve better performance compared with various strong baselines in terms of multiple ROUGE metrics.

N-GPETS: Neural Attention Graph-Based Pretrained Statistical Model for Extractive Text Summarization.

The extractive summarization approach involves selecting the source document's salient sentences to build a summary. One of the most important aspects of extractive summarization is learning and modelling cross-sentence associations. Inspired by the popularity of Transformer-based Bidirectional Encoder Representations (BERT) pretrained linguistic model and graph attention network (GAT) having a sophisticated network that captures intersentence associations, this research work proposes a novel neural model N-GPETS by combining heterogeneous graph attention network with BERT model along with statistical approach using TF-IDF values for extractive summarization task. Apart from sentence nodes, N-GPETS also works with different semantic word nodes of varying granularity levels that serve as a link between sentences, improving intersentence interaction. Furthermore, proposed N-GPETS becomes more improved and feature-rich by integrating graph layer with BERT encoder at graph initialization step rather than employing other neural network encoders such as CNN or LSTM. To the best of our knowledge, this work is the first attempt to combine the BERT encoder and TF-IDF values of the entire document with a heterogeneous attention graph structure for the extractive summarization task. The empirical outcomes on benchmark news data sets CNN/DM show that the proposed model N-GPETS gets favorable results in comparison with other heterogeneous graph structures employing the BERT model and graph structures without the BERT model.

Template-based Abstractive Microblog Opinion Summarization

AbstractWe introduce the task of microblog opinion summarization (MOS) and share a dataset of 3100 gold-standard opinion summaries to facilitate research in this domain. The dataset contains summaries of tweets spanning a 2-year period and covers more topics than any other public Twitter summarization dataset. Summaries are abstractive in nature and have been created by journalists skilled in summarizing news articles following a template separating factual information (main story) from author opinions. Our method differs from previous work on generating gold-standard summaries from social media, which usually involves selecting representative posts and thus favors extractive summarization models. To showcase the dataset’s utility and challenges, we benchmark a range of abstractive and extractive state-of-the-art summarization models and achieve good performance, with the former outperforming the latter. We also show that fine-tuning is necessary to improve performance and investigate the benefits of using different sample sizes.

Unsupervised abstractive summarization via sentence rewriting

Unsupervised extractive summarization aims to extract salient sentences from the document without labeled corpus. Existing methods have achieved promising progress, thanks to the power of large-scale pre-trained language models and high-quality contextualized representations. However, extractive summaries often fail to maintain smooth transitions between sentences and struggle to form a coherent and fluent text due to splicing of sentences. Nevertheless, to the best of our knowledge, very few studies currently focus on unsupervised abstractive summarization. Inspired by the intuitive human process of writing summaries, which involves extracting salient sentences first and then reconstructing them, in this paper, we propose an Extract-then-Abstract framework to generate more coherent and human-like summary. Specifically, we first adopt extractive summarization model as summarizer to generate extractive summary in the extraction stage. Then in the abstraction stage, we propose a BART-based sentence write model to generate more coherent and fluent abstractive summary. To this end, we design a novel parallel data creation method for our rewrite model by proposing an effective sentence sampling strategy without any manual annotation cost. Extensive experiments including automatic evaluation and human evaluation demonstrate that our framework consistently outperforms strong baselines for unsupervised abstractive summarization and can generate more coherent and human-like summary while maintaining in competitive ROUGE scores for unsupervised extractive summarization.

HetTreeSum: A Heterogeneous Tree Structure-based Extractive Summarization Model for Scientific Papers

Scientific paper summarization aims at generating a short and concise digest while preserving important information of the original document. Currently, scientific paper summarization faces two main challenges. First, inter-sentence relations are hard to learn, especially in the case of long-form scientific papers. Second, structural information of the well-structured scientific papers has not been fully exploited. To overcome the above two challenges, we propose a novel Heterogeneous Tree structure-based extractive Summarization (HetTreSum) model, where each document is modeled as a tree structure to learn inter-sentence relations and structural information of the original document is incorporated, enabling the tree structure to have a global perspective of the whole document. Then an iterative updating strategy is presented to interactively refine nodes of the tree structure for better contextualized representations, which can further enhance summarization performance. Experimental results on PubMed and arXiv datasets show that our proposed HetTreeSum model achieves significantly advanced performance compared with various scientific paper summarization models.

T-BERTSum: Topic-Aware Text Summarization Based on BERT

In the era of social networks, the rapid growth of data mining in information retrieval and natural language processing makes automatic text summarization necessary. Currently, pretrained word embedding and sequence to sequence models can be effectively adapted in social network summarization to extract significant information with strong encoding capability. However, how to tackle the long text dependence and utilize the latent topic mapping has become an increasingly crucial challenge for these models. In this article, we propose a topic-aware extractive and abstractive summarization model named T-BERTSum, based on Bidirectional Encoder Representations from Transformers (BERTs). This is an improvement over previous models, in which the proposed approach can simultaneously infer topics and generate summarization from social texts. First, the encoded latent topic representation, through the neural topic model (NTM), is matched with the embedded representation of BERT, to guide the generation with the topic. Second, the long-term dependencies are learned through the transformer network to jointly explore topic inference and text summarization in an end-to-end manner. Third, the long short-term memory (LSTM) network layers are stacked on the extractive model to capture sequence timing information, and the effective information is further filtered on the abstractive model through a gated network. In addition, a two-stage extractive–abstractive model is constructed to share the information. Compared with the previous work, the proposed model T-BERTSum focuses on pretrained external knowledge and topic mining to capture more accurate contextual representations. Experimental results on the CNN/Daily mail and XSum datasets demonstrate that our proposed model achieves new state-of-the-art results while generating consistent topics compared with the most advanced method.

STV-BEATS: Skip Thought Vector and Bi-Encoder based Automatic Text Summarizer

A novel text summarization framework referred to as Skip-Though Vector and Bi-encoder Based Automatic Text Summarization (STV–BEATS) is proposed in this paper. STV–BEATS utilizes — (a) skip-though vector to generate sentence-based embedding; and (b) Long Short-Term Memory (LSTM) based deep autoencoder to reduce dimensions of skip thought vectors. STV–BEATS works in the conjunction of extractive and abstractive summarization models to enhance the overall quality of the results. For each sentence, relevance and novelty metrics are calculated on the intermediate representation of the deep autoencoder to generate the final sentence score. The highly scored sentences are selected to generate an extractive summary. On the other hand, the abstractive part is composed of two encoders and a decoder which works as — (a) the first GRU-based bi-directional encoder and decoder work as basic sequence-to-sequence model on the extractive summary; and (b) the second GRU-based unidirectional encoder is used for fine encoding. Extensive computer experiments are conducted to determine the effectiveness of the STV–BEATS. Three standard benchmark datasets, namely, CNN/Daily Mail, DUC-2004, and DUC-2002 are used during experiments. Further, recall-oriented understudy for gisting evaluation (ROUGE) is used for validation of the STV–BEATS. Result reveals that the proposed STV–BEATS is capable of effective text summarization and achieves substantially better results over the state-of-the-art models.

Extractive text summarization model based on advantage actor-critic and graph matrix methodology.

The automatic text summarization task faces great challenges. The main issue in the area is to identify the most informative segments in the input text. Establishing an effective evaluation mechanism has also been identified as a major challenge in the area. Currently, the mainstream solution is to use deep learning for training. However, a serious exposure bias in training prevents them from achieving better results. Therefore, this paper introduces an extractive text summarization model based on a graph matrix and advantage actor-critic (GA2C) method. The articles were pre-processed to generate a graph matrix. Based on the states provided by the graph matrix, the decision-making network made decisions and sent the results to the evaluation network for scoring. The evaluation network got the decision results of the decision-making network and then scored them. The decision-making network modified the probability of the action based on the scores of the evaluation network. Specifically, compared with the baseline reinforcement learning-based extractive summarization (Refresh) model, experimental results on the CNN/Daily Mail dataset showed that the GA2C model led on Rouge-1, Rouge-2 and Rouge-A by 0.70, 9.01 and 2.73, respectively. Moreover, we conducted multiple ablation experiments to verify the GA2C model from different perspectives. Different activation functions and evaluation networks were used in the GA2C model to obtain the best activation function and evaluation network. Two different reward functions (Set fixed reward value for accumulation (ADD), Rouge) and two different similarity matrices (cosine, Jaccard) were combined for the experiments.

StarSum: A Star Architecture Based Model for Extractive Summarization

Extractive summarization aims to produce a concise summary while retaining the key information through the way of selecting sentences from the original document. Under such background, learning inter-sentence relations has hitherto been the issue of most concern. In this study, we propose a Star architecture based model for extractive summarization (StarSum), that takes advantage of self-attention strategy based Transformer and star-shaped structure, models sentences within a document as satellite nodes and introduces a virtual star node, constructs a star model for each document to learn inter-sentence relations. Based on the constructed star-shaped model, we further develop two sentence representation learning algorithms, namely star guiding satellite (SGS) algorithm and star incorporating satellite (SIS) algorithm, in order to extract summary-worthy sentences. Experimental results on CNN/Daily Mail, New York Times (NYT) and XSum datasets prove that StarSum model achieves advanced performance for extractive summarization and has comparable performance to the state-of-the-art extractive summarization model. The results also demonstrate that the SIS algorithm is more effective than the SGS algorithm.