Summary Sentences Research Articles

HindiSumm: A Hindi Abstractive Summarization benchmark dataset

Abstractive Text Summarization (ATS) is a task to create a novel summary by generating fresh sentences incorporating new words or rephrasing the article. It is a complex task as the model needs to understand the semantic similarity between the sentences of the text. To fulfill this, there is a need for a large annotated benchmark dataset, which is available for resource-rich languages such as English and non-indic languages. In contrast, for the less-resourced languages, such as Indic languages, the available datasets are limited and involve very short summary sentences. Hence, a language-specific abstractive summarization dataset called HindiSumm was introduced for Hindi, consisting of 570,000 text-summary pairs from Navbharat Times across 21 domains. The HindiSumm dataset’s efficiency is evaluated extrinsically and intrinsically by using various metrics. Furthermore, two recent multilingual-cased pre-trained models are fine-tuned on the HindiSumm dataset individually. In addition, an ensembled approach using weighted averaging is also incorporated to check the efficacy of the proposed dataset. The model is tested with the in-house created dataset, and results are evaluated on ROUGE scores and show significant improvements of around 13.2% for the proposed HindiSumm compared to other benchmark datasets. In the future, the HindiSumm dataset will promote the progress of ATS for the Indian language.

A design of computational fuzzy set-based semantics for extracting linguistic summaries

Linguistic summarization is to extract a set of summary sentences in the form of natural language, so-called linguistic summaries, from numeric data. The extracted linguistic summaries should be compact and diverse, and have a validity measure greater than a given threshold, so genetic algorithms are applied to extract such linguistic summaries. Besides, the interpretability of linguistic summary content is considered in recent studies in such a way that enlarged hedge algebras are applied to generate multi-semantic structures for linguistic words of linguistic variables ensuring the interpretability of the content of the linguistic summaries. However, the membership function of computational fuzzy-set-based semantics of linguistic words is usually in the shape of trapezoid. In this paper, a membership function of the form S function is applied to improve the quality of extracted linguistic summaries. Besides, the applied algorithm is parallelized to reduce running time. The experimental results with the creep dataset have demonstrated the effectiveness of the proposed method

Helminth-derived metabolites induce tolerogenic functional, metabolic, and transcriptional signatures in dendritic cells that attenuate experimental colitis.

Inflammatory bowel diseases (IBD) are chronic inflammatory diseases in which abdominal pain, bloody diarrhea, weight loss, and fatigue collectively result in diminished quality of patient life. The disappearance of intestinal helminth infections in Western societies is associated with an increased prevalence of IBD and other immune-mediated inflammatory diseases. Evidence indicates that helminths induce tolerogenic dendritic cells (tolDCs), which promote intestinal tolerance and attenuate intestinal inflammation characteristic of IBD, but the exact mechanism is unclear. Helminth-derived excretory-secretory (HES) products including macromolecules, proteins, and polysaccharides have been shown to modulate the antigen presenting function of DCs with down-stream effects on effector CD4 + T cells. Previous studies indicate that DCs in helminth-infected animals induce tolerance to unrelated antigens and DCs exposed to HES display phenotypic and functional features of tolDCs. Here, we identify that nonpolar metabolites (HnpM) produced by a helminth, the murine gastrointestinal nematode Heligmosomoides polygyrus bakeri (Hpb), induce tolDCs as evidenced by decreased LPS-induced TNF and increased IL-10 secretion and reduced expression of MHC-II, CD86, and CD40. Furthermore, these DCs inhibited OVA-specific CD4 + T cell proliferation and induced CD4 + Foxp3 + regulatory T cells. Adoptive transfer of HnpM-induced tolDCs attenuated DSS-induced intestinal inflammation characteristic of IBD. Mechanistically, HnpM induced metabolic and transcriptional signatures in BMDCs consistent with tolDCs. Collectively, our findings provide groundwork for further investigation into novel mechanisms regulating DC tolerance and the role of helminth secreted metabolites in attenuating intestinal inflammation associated with IBD. Summary Sentence: Metabolites produced by Heligmosomoides polygyrus induce metabolic and transcriptional changes in DCs consistent with tolDCs, and adoptive transfer of these DCs attenuated DSS-induced intestinal inflammation.

Protease expression in the human and rat cumulus-oocyte complex during the periovulatory period: a role in cumulus-oocyte complex migration†.

The migratory and matrix-invading capacities of the cumulus-oocyte complex have been shown to be important for the ovulatory process. In metastatic cancers, these capacities are due to increased expression of proteases, however, there is limited information on protease expression in the cumulus-oocyte complexes. The present study examined cumulus-oocyte complex expression of plasmins, matrix metalloproteases, and A Disintegrin and Metalloproteinase with Thrombospondin Motifs family members in the rat and human. In the rat, human chorionic gonadotropin (hCG) administration increased cumulus-oocyte complex expression of Mmp2, Mmp9, Mmp13, Mmp14, Mmp16, Adamts1, and the protease inhibitors Timp1, Timp3, and Serpine1 by 8-12h. This ovulatory induction of proteases in vivo could be mimicked by forskolin and ampiregulin treatment of cultured rat cumulus-oocyte complexes with increases observed in Mmp2, Mmp13, Mmp14, Mmp16, Mmp19, Plat, and the protease inhibitors Timp1, Timp3, and Serpine1. Comparison of expression between rat cumulus-oocyte complexes and granulosa cells at the time of ovulation showed decreased Mmp9 and increased Mmp13, Mmp14, Mmp16, Adamts1, Timp1, and Timp3 expression in the cumulus-oocyte complexes. In human, comparison of expression between cumulus and granulosa cells at the time of in vitro fertilization retrieval showed decreased MMP1, MMP2, MMP9, and ADAMTS1, while expression of MMP16, TIMP1, and TIMP3 were increased. Treatment of expanding rat cumulus-oocyte complexes with a broad spectrum matrix metalloproteases inhibitor, GM6001, significantly reduced the migration of cumulus cells in vitro. These data provide evidence that multiple proteases and their inhibitors are expressed in the cumulus-oocyte complex and play an important role in imparting the migratory phenotype of the cumulus-oocyte complex at the time of ovulation. Summary Sentence Multiple proteases and their inhibitors are induced in the cumulus-oocyte complex (COC) during the periovulatory period and potentially play an important role in imparting the migratory phenotype of the COC at the time of ovulation.

Leveraging artificial intelligence to summarize abstracts in lay language for increasing research accessibility and transparency.

Returning aggregate study results is an important ethical responsibility to promote trust and inform decision making, but the practice of providing results to a lay audience is not widely adopted. Barriers include significant cost and time required to develop lay summaries and scarce infrastructure necessary for returning them to the public. Our study aims to generate, evaluate, and implement ChatGPT 4 lay summaries of scientific abstracts on a national clinical study recruitment platform, ResearchMatch, to facilitate timely and cost-effective return of study results at scale. We engineered prompts to summarize abstracts at a literacy level accessible to the public, prioritizing succinctness, clarity, and practical relevance. Researchers and volunteers assessed ChatGPT-generated lay summaries across five dimensions: accuracy, relevance, accessibility, transparency, and harmfulness. We used precision analysis and adaptive random sampling to determine the optimal number of summaries for evaluation, ensuring high statistical precision. ChatGPT achieved 95.9% (95% CI, 92.1-97.9) accuracy and 96.2% (92.4-98.1) relevance across 192 summary sentences from 33 abstracts based on researcher review. 85.3% (69.9-93.6) of 34 volunteers perceived ChatGPT-generated summaries as more accessible and 73.5% (56.9-85.4) more transparent than the original abstract. None of the summaries were deemed harmful. We expanded ResearchMatch's technical infrastructure to automatically generate and display lay summaries for over 750 published studies that resulted from the platform's recruitment mechanism. Implementing AI-generated lay summaries on ResearchMatch demonstrates the potential of a scalable framework generalizable to broader platforms for enhancing research accessibility and transparency.

Automatic summarization model based on clustering algorithm

Extractive document summary is usually seen as a sequence labeling task, which the summary is formulated by sentences from the original document. However, the selected sentences usually are high redundancy in semantic space, so that the composed summary are high semantic redundancy. To alleviate this problem, we propose a model to reduce the semantic redundancy of summary by introducing the cluster algorithm to select difference sentences in semantic space and we improve the base BERT to score sentences. We evaluate our model and perform significance testing using ROUGE on the CNN/DailyMail datasets compare with six baselines, which include two traditional methods and four state-of-art deep learning model. The results validate the effectiveness of our approach, which leverages K-means algorithm to produce more accurate and less repeat sentences in semantic summaries.

Reproductive & developmental toxicity of quaternary ammonium compounds†.

Quaternary ammonium compounds are a class of chemicals commonly used as disinfectants in household and healthcare settings. Their usage has significantly increased in recent years due to the COVID-19 pandemic. In addition, quaternary ammonium compounds have replaced the recently banned disinfectants triclosan and triclocarban in consumer products. Quaternary ammonium compounds are found in daily antimicrobial and personal care products such as household disinfectants, mouthwash, and hair care products. Due to the pervasiveness of quaternary ammonium compounds in daily use products, humans are constantly exposed. However, little is known about the health effects of everyday quaternary ammonium compound exposure, particularly effects on human reproduction and development. Studies that investigate the harmful effects of quaternary ammonium compounds on reproduction are largely limited to high-dose studies, which may not be predictive of low-dose, daily exposure, especially as quaternary ammonium compounds may be endocrine-disrupting chemicals. This review analyzes recent studies on quaternary ammonium compound effects on reproductive health, identifies knowledge gaps, and recommends future directions in quaternary ammonium compound-related research. Summary Sentence Quaternary ammonium compounds, a class of disinfecting compounds that have skyrocketed in usage during the COVID-19 pandemic, are emerging as reproductive and developmental toxicants.

METTL14 mediates nerve growth factor-stimulated testosterone synthesis in porcine theca cells†.

Ovarian theca cells produce testosterone, which acts as a vital precursor substance for synthesizing estrogens during follicular development. Nerve growth factor (NGF) has been shown to participate in reproductive physiology, specifically to follicular development and ovulation. There is currently no available data on the impact of NGF on testosterone synthesis in porcine theca cells. Furthermore, m6A modification is the most common internal modification in eukaryotic mRNAs that are closely associated with female gametogenesis, follicle development, ovulation, and other related processes. It is also uncertain whether the three main enzymes associated with m6A, such as Writers, Erasers, and Readers, play a role in this process. The present study, with an in vitro culture model, investigated the effect of NGF on testosterone synthesis in porcine theca cells and the role of Writers-METTL14 in this process. It was found that NGF activates the PI3K/AKT signaling pathway through METTL14, which regulates testosterone synthesis in porcine theca cells. This study will help to further elucidate the mechanisms by which NGF regulates follicular development and provide new therapeutic targets for ovary-related diseases in female animals. Summary Sentence The present study investigated the effect of NGF on testosterone synthesis in porcine theca cells. It was found that NGF activates the PI3K/AKT signaling pathway through METTL14, which regulates testosterone synthesis in porcine theca cells.

Increasing Interest in Learning by Applying Learning Summary Writing Skills Text Content Using Inspiring Stories Important Video Media for Student Class IX

The skill of writing a summary of the contents of an inspirational story text is very important for students to simplify and clarify something too long. The higher the student's writing ability, the higher the level of understanding of what has been conveyed/listened to. However, most students lack mastery of summarizing. This research aims to increase students' interest in learning by implementing learning skills in writing summaries of the contents of inspirational story texts using illustrated video media in class IX Plus students at Al-Azhar Medan Middle School, Medan Johor District, academic year 2023-2024. The subjects of this research were class IX Plus students at Al-Azhar Medan Middle School, Medan Johor District, for the 2023-2024 academic year, consisting of 40 students. The sample in this study is defined as individuals/students who constitute the entire population studied. The method used in this research is classroom action research (PTK). The indicators for this research are teachers who teach the material to write summary sentences and paragraphs on the contents of inspirational story texts. The data analysis technique used was classroom action research through Cycle I and Cycle II. Through the stages: class action plan, implementation, observation, and reflection according to the learning implementation plan (RPP) and syllabus applicable in the learning curriculum at Al-Azhar Middle School Medan. Based on data from 40 students, it was found that the percentage of classical student learning completeness in cycle I was 75% of the number of students who completed it, while 10 students did not complete it, and in cycle II it was 95% of the number of students who completed it, there were 38 students who did not complete it, while 2 did not complete it.

Deep Learning Based Hybrid Technique for Improved Summarization of Abstractive Text

Аbstrасt: From various source sentences, construct summary sentences by merging facts. The process of preserving information material and overall meaning while condensing it into a shorter representation is known as abstractive text summarization, or ATS. It takes a lot of effort and time for humans to manually summarise large textual publications. In this paper, we present an ATS framework (ATSDL) based on (Long Short-Term Memory-Convolutional Neural Network) LSTM- CNN that can generate new sentences by investigating semantic phrases, which are finer- grained fragments than sentences. ATSDL, in contrast to current abstraction- based methods, consists of two primary phases: the first extracts phrases from source sentences, and the second uses deep learning to produce text summaries. Experimental results on CNN and Daily Mail datasets show that our ATSDL framework outperforms the state-of-the-art models in terms of both syntactic structure and semantics, and achieves competitive results on manual linguistic quality evaluation. In this application hybrid model is giving better performance over other state of art techniques.

Shear wave elastography to assess stiffness of the human ovary and other reproductive tissues across the reproductive lifespan in health and disease†.

The ovary is one of the first organs to show overt signs of aging in the human body, and ovarian aging is associated with a loss of gamete quality and quantity. The age-dependent decline in ovarian function contributes to infertility and an altered endocrine milieu, which has ramifications for overall health. The aging ovarian microenvironment becomes fibro-inflammatory and stiff with age, and this has implications for ovarian physiology and pathology, including follicle growth, gamete quality, ovulation dynamics, and ovarian cancer. Thus, developing a non-invasive tool to measure and monitor the stiffness of the human ovary would represent a major advance for female reproductive health and longevity. Shear wave elastography is a quantitative ultrasound imaging method for evaluation of soft tissue stiffness. Shear wave elastography has been used clinically in assessment of liver fibrosis and characterization of tendinopathies and various neoplasms in thyroid, breast, prostate, and lymph nodes as a non-invasive diagnostic and prognostic tool. In this study, we review the underlying principles of shear wave elastography and its current clinical uses outside the reproductive tract as well as its successful application of shear wave elastography to reproductive tissues, including the uterus and cervix. We also describe an emerging use of this technology in evaluation of human ovarian stiffness via transvaginal ultrasound. Establishing ovarian stiffness as a clinical biomarker of ovarian aging may have implications for predicting the ovarian reserve and outcomes of Assisted Reproductive Technologies as well as for the assessment of the efficacy of emerging therapeutics to extend reproductive longevity. This parameter may also have broad relevance in other conditions where ovarian stiffness and fibrosis may be implicated, such as polycystic ovarian syndrome, late off target effects of chemotherapy and radiation, premature ovarian insufficiency, conditions of differences of sexual development, and ovarian cancer. Summary sentence: Shear Wave Elastography is a non-invasive technique to study human tissue stiffness, and here we review its clinical applications and implications for reproductive health and disease.

Role of miRNAs in glucose metabolism of mouse cumulus cells†.

It is known that the oocyte has a limited capacity to acquire and metabolize glucose, and it must rely on cumulus cells (CCs) to take up glucose and produce pyruvate for use to produce ATP through oxidative phosphorylation. We therefore propose that miRNAs might regulate glucose metabolism (GM) in CCs and might be used as markers for oocyte quality assessment. Here, mouse CC models with impaired glycolysis or pentose phosphate pathway (PPP) were established, and miRNAs targeting the key enzymes in glycolysis/PPP were predicted using the miRNA target prediction databases. Expression of the predicted miRNAs was compared between CCs with normal and impaired glycolysis/PPP to identify candidate miRNAs. Function of the candidate miRNAs was validated by transfecting CCs or cumulus-oocyte-complexes (COCs) with miRNA inhibitors and observing effects on glucose metabolites of CCs and on competence of oocytes. The results validated that miR-23b-3p, let-7b-5p, 34b-5p and 145a-5p inhibited glycolysis, and miR-24-3p, 3078-3p,183-5p and 7001-5p inhibited PPP of CCs. Our observation using a more physiologically relevant model (intact cultured COCs) further validated the four glycolysis-targeting miRNAs we identified. Furthermore, miR-let-7b-5p, 34b-5p and 145a-5p may also inhibit PPP, as they decreased the production of glucose-6-phosphate. In conclusion, miRNAs play critical roles in GM of CCs and may be used as markers for oocyte quality assessment. Summary sentence: We identified and validated eight new miRNAs that inhibit glycolysis and/or pentose phosphate pathways in cumulus cells (CCs) suggesting that miRNAs play critical roles in glucose metabolism of CCs and may be used for oocyte quality markers.

Joint inter-word and inter-sentence multi-relation modeling for summary-based recommender system

Review is an essential piece of information that influences users’ decisions, but excessively long reviews not only degrade the user experience but also affect the accuracy of the recommender system. Therefore, Joint Inter-Word and Inter-Sentence Multi-Relation Modeling for the Summary-based Recommender System (MRSR) is proposed in this paper. In MRSR, the concise summary information serves as representation data, and a multi-relation modeling module is designed to construct user and item characteristics from two levels. Specifically, the inter-word relation modeling module, which consists of the Transformer and the pooling layer, is used to learn the long dependencies of summaries and extract word-level features by calculating the relative weights between words within sentences. The inter-sentence relation modeling module is used to enrich the sentence-level features of users and items, where an attention mechanism is employed to perceive the relative weights between different summary sentences. Finally, the fusion layer based on multi-head attention and the prediction layer based on attentional factorization machine are implemented to conduct the shallow and deep interactions between user and item features, based on which MRSR completes the final rating prediction. Extensive experimental results on five publicly available datasets demonstrate that MRSR achieves a 5.94% improvement in RMSE metrics compared to state-of-the-art methods. Furthermore, the accuracy of most existing models is improved by 1%∼2% while the inference time is reduced by 10% by utilizing summaries as representation data. It proves the efficiency and effectiveness of our proposed approach, which has promising application prospects.

Keyword Extraction-based Library Intelligence Services: Challenges, Adaptations and Reinvention

Abstract This paper proposes a domain knowledge base construction platform for thematic intelligence services of library intelligence institutions. The platform proposes a keyword extraction algorithm applicable to the literature resources of library intelligence services based on a statistical model and N-Gram model combined with feature word information. The automatic summary algorithm based on statistics combines the information of sentence content and location to comprehensively evaluate the weights of candidate summary sentences and achieve automatic summary extraction. Based on the HDP topic model, a literature recommendation model oriented to the resource retrieval context is proposed, which can convert the literature feature similarity calculation to the topic similarity calculation with implied literature feature semantics from the perspective of semantic analysis and improve the accuracy of recommended literature. The technical feasibility of this system platform is verified through relevant data experiments. Regarding the accuracy of the literature recommendation, the platform of this paper and the traditional method are 86.36% and 75.16%, respectively, and the recall rate is 67.29% and 57.25%, respectively. The platform scored 3.85 on the comprehensive evaluation of library intelligence service, close to the excellent level. Therefore, the platform can realize various functions, from resource collection and organization to resource processing and utilization, with certain practical significance.

Summarization of Lengthy Legal Documents via Abstractive Dataset Building: An Extract-then-Assign Approach

Development of effective automatic summarization approaches for legal documents suffer from several challenges like extremely long document-summary pairs, lack of large scale training datasets with tractable document-summary token lengths, etc. In this work, we deal with the problem of legal document summarization by building a modified abstractive dataset from the original dataset. This ensures that the length of each document-summary pair is manageable and can be processed by the state of the art summarization approaches (such as BART). Secondly, we deal with the data scarcity problem by creating more number of training samples, from each of the original document-summary pair. This is done by creating multiple extractive summaries from each sample in the original dataset, following which ground-truth summary sentences are assigned to each of the extractive summary to generate new training samples. This results in a larger training dataset that can be utilized for fine-tuning summarization models. Our proposed approach has been evaluated on two different legal datasets- BillSum and Forum of Information Retrieval Evaluation (FIRE). With respect to the ROUGE metrics, the proposed approach is able to outperform pre-trained BART model fine-tuned on original dataset by (3−8)% for FIRE test sets, and by (1−3)% for the BillSum test sets. Considering the BERTScore metrics, the proposed approach obtains (1−2)% improvements on the FIRE test sets, while for the BillSum test sets (3−8)% improvements are observed. Such improvements suggest that the proposed dataset building approach can help achieve improved abstractive summarization of lengthy legal documents.

Effective News Text Summarization Techniques

n proposed work, we successfully implemented a news text summarization system using Natural Language Processing (NLP) techniques and the Latent Semantic Analysis (LSA) algorithm. The purpose of our project was to extract important information from a large volume of news articles and present it in a concise and easily understandable manner. To achieve this, we utilized the LSA algorithm, which is known for its ability to capture the underlying semantic structure of text. LSA employs a mathematical model to analyse relationships between words in a document, creating a semantic representation where words with similar contexts are grouped together in a vector space. The LSA-based summarization process involved several steps. First, we pre-processed the news articles by removing stop words, punctuation, and other non-relevant elements. Then, we constructed a term-document matrix, where rows represented words and columns represented documents, with matrix values representing word frequencies. Next, we applied Singular Value Decomposition (SVD) to the term-document matrix. SVD helped reduce the matrix's dimensionality by identifying the most important latent semantic concepts. This resulted in a lower-dimensional representation that captured the essential information. Finally, we identified the most important sentences in the news articles by measuring the cosine similarity between each sentence and the summary. Sentences with the highest cosine similarity scores were selected as summary sentences. The proposed system demonstrated the effectiveness of the LSA algorithm for news text summarization. By capturing the semantic structure of the text, it generated summaries that allowed users to understand the key points of a news article quickly and easily. Our implementation had practical applications for content recommendation systems, news aggregation platforms, and personalized news feeds. However, it is important to acknowledge the limitations of the LSA algorithm. It may struggle with handling idiomatic expressions and can be sensitive to the quality of the input data. These considerations highlight the need for ongoing research and development to enhance the performance and robustness of news text summarization systems.

Generation of a novel Stra8-driven Cre recombinase strain for use in pre-meiotic germ cells in mice†.

The development of oocytes occurs over a broad time frame, starting at the earliest stages of embryogenesis and continuing into adulthood. Conditional knockout technologies such as the Cre/loxP recombination system are useful for analyzing oocyte development at specific stages, but not every time frame has appropriate Cre drivers, for instance, during oocyte meiotic initiation through early prophase I in the embryo. Here, we generated a novel knockin mouse line that produces a bicistronic transcript from the endogenous Stra8 locus that includes a "self-cleaving" 2A peptide upstream of cre. This allows for high efficiency cleavage and production of both proteins individually and results in expression of cre in both male and female gonads at the biologically relevant stage. Fluorescent reporter analysis confirms that this line recapitulates endogenous Stra8 expression in both sexes and does not affect fertility of heterozygous nor homozygous mice. This line, named Stra8P2Acre, adds to the repertoire of germ-cell specific cre driver lines and, importantly, allows for deletion of target genes during key embryonic oocyte developmental stages, including early events in meiosis. Summary Sentence Generation of a novel cre recombinase knockin to the Stra8 locus allows production of Stra8 and cre without affecting fertility.

Joint learning of text alignment and abstractive summarization for long documents via unbalanced optimal transport

Abstract Recently, neural abstractive text summarization (NATS) models based on sequence-to-sequence architecture have drawn a lot of attention. Real-world texts that need to be summarized range from short news with dozens of words to long reports with thousands of words. However, most existing NATS models are not good at summarizing long documents, due to the inherent limitations of their underlying neural architectures. In this paper, we focus on the task of long document summarization (LDS). Based on the inherent section structures of source documents, we divide an abstractive LDS problem into several smaller-sized problems. In this circumstance, how to provide a less-biased target summary as the supervision for each section is vital for the model’s performance. As a preliminary, we formally describe the section-to-summary-sentence (S2SS) alignment for LDS. Based on this, we propose a novel NATS framework for the LDS task. Our framework is built based on the theory of unbalanced optimal transport (UOT), and it is named as UOTSumm. It jointly learns three targets in a unified training objective, including the optimal S2SS alignment, a section-level NATS summarizer, and the number of aligned summary sentences for each section. In this way, UOTSumm directly learns the text alignment from summarization data, without resorting to any biased tool such as ROUGE. UOTSumm can be easily adapted to most existing NATS models. And we implement two versions of UOTSumm, with and without the pretrain-finetune technique. We evaluate UOTSumm on three publicly available LDS benchmarks: PubMed, arXiv, and GovReport. UOTSumm obviously outperforms its counterparts that use ROUGE for the text alignment. When combined with UOTSumm, the performance of two vanilla NATS models improves by a large margin. Besides, UOTSumm achieves better or comparable performance when compared with some recent strong baselines.

Automatic Summarization for Academic Articles using Deep Learning and Reinforcement Learning with Viewpoints

The purpose of this research is to develop a Viewpoint Refinement in Automatic Summarization (VPRAS) system for research articles. The system will reflect viewpoints of survey to support surveys stage for researchers and students. We collect academic articles using web scraping technology and construct training data by combining sections and sentences through analysis of the article's PDF structure. We use machine learning techniques to classify sentences in Japanese articles into viewpoints. In addition to supervised learning, we introduce reinforcement learning and Dynamic Programming (DP) to extract important sentences for each viewpoint. Finally, we implemented an agent to automatically extract summary sentences based on a reward function.

Attentional Extractive Summarization

In this work, a general theoretical framework for extractive summarization is proposed—the Attentional Extractive Summarization framework. Although abstractive approaches are generally used in text summarization today, extractive methods can be especially suitable for some applications, and they can help with other tasks such as Text Classification, Question Answering, and Information Extraction. The proposed approach is based on the interpretation of the attention mechanisms of hierarchical neural networks, which compute document-level representations of documents and summaries from sentence-level representations, which, in turn, are computed from word-level representations. The models proposed under this framework are able to automatically learn relationships among document and summary sentences, without requiring Oracle systems to compute the reference labels for each sentence before the training phase. These relationships are obtained as a result of a binary classification process, the goal of which is to distinguish correct summaries for documents. Two different systems, formalized under the proposed framework, were evaluated on the CNN/DailyMail and the NewsRoom corpora, which are some of the reference corpora in the most relevant works on text summarization. The results obtained during the evaluation support the adequacy of our proposal and suggest that there is still room for the improvement of our attentional framework.