• Title/Summary/Keyword: Rouge

Search Result 78, Processing Time 0.025 seconds

A Study on the Self-Awareness of Infants (영아기 자아인식 능력에 관한 연구)

  • Kim, Yang Eun;Cho, Bok Hee
    • Korean Journal of Child Studies
    • /
    • v.23 no.2
    • /
    • pp.17-30
    • /
    • 2002
  • The subjects of this study on self-awareness were sixty-four 8- to 30-month-old infants. They were tested with the "Behaviors Observed in Mirror Conditions" (Lewis & Brooks-Gunn, 1979; translated by Jong Moo Kim, 1989). Analyses were by descriptive statistics and ANOVA. Most infants responded when they saw their reflection in the mirror, and their self-awareness increased with age. Most infants under 17 months didn't respond before and after rouge was painted on their noses, but by 18- to 23-months, they started to show a response. Most of the infants older than 24 months noticed something unusual and tried to rub the rouge off when they saw their reflection. This study provides evidence that infants begin to show self-awareness after 18 months of age, and this self-awareness is well developed after 24 months of age.

  • PDF

Improving Abstractive Summarization by Training Masked Out-of-Vocabulary Words

  • Lee, Tae-Seok;Lee, Hyun-Young;Kang, Seung-Shik
    • Journal of Information Processing Systems
    • /
    • v.18 no.3
    • /
    • pp.344-358
    • /
    • 2022
  • Text summarization is the task of producing a shorter version of a long document while accurately preserving the main contents of the original text. Abstractive summarization generates novel words and phrases using a language generation method through text transformation and prior-embedded word information. However, newly coined words or out-of-vocabulary words decrease the performance of automatic summarization because they are not pre-trained in the machine learning process. In this study, we demonstrated an improvement in summarization quality through the contextualized embedding of BERT with out-of-vocabulary masking. In addition, explicitly providing precise pointing and an optional copy instruction along with BERT embedding, we achieved an increased accuracy than the baseline model. The recall-based word-generation metric ROUGE-1 score was 55.11 and the word-order-based ROUGE-L score was 39.65.

Empirical Study for Automatic Evaluation of Abstractive Summarization by Error-Types (오류 유형에 따른 생성요약 모델의 본문-요약문 간 요약 성능평가 비교)

  • Seungsoo Lee;Sangwoo Kang
    • Korean Journal of Cognitive Science
    • /
    • v.34 no.3
    • /
    • pp.197-226
    • /
    • 2023
  • Generative Text Summarization is one of the Natural Language Processing tasks. It generates a short abbreviated summary while preserving the content of the long text. ROUGE is a widely used lexical-overlap based metric for text summarization models in generative summarization benchmarks. Although it shows very high performance, the studies report that 30% of the generated summary and the text are still inconsistent. This paper proposes a methodology for evaluating the performance of the summary model without using the correct summary. AggreFACT is a human-annotated dataset that classifies the types of errors in neural text summarization models. Among all the test candidates, the two cases, generation summary, and when errors occurred throughout the summary showed the highest correlation results. We observed that the proposed evaluation score showed a high correlation with models finetuned with BART and PEGASUS, which is pretrained with a large-scale Transformer structure.

Automatic Text Summarization based on Selective Copy mechanism against for Addressing OOV (미등록 어휘에 대한 선택적 복사를 적용한 문서 자동요약)

  • Lee, Tae-Seok;Seon, Choong-Nyoung;Jung, Youngim;Kang, Seung-Shik
    • Smart Media Journal
    • /
    • v.8 no.2
    • /
    • pp.58-65
    • /
    • 2019
  • Automatic text summarization is a process of shortening a text document by either extraction or abstraction. The abstraction approach inspired by deep learning methods scaling to a large amount of document is applied in recent work. Abstractive text summarization involves utilizing pre-generated word embedding information. Low-frequent but salient words such as terminologies are seldom included to dictionaries, that are so called, out-of-vocabulary(OOV) problems. OOV deteriorates the performance of Encoder-Decoder model in neural network. In order to address OOV words in abstractive text summarization, we propose a copy mechanism to facilitate copying new words in the target document and generating summary sentences. Different from the previous studies, the proposed approach combines accurate pointing information and selective copy mechanism based on bidirectional RNN and bidirectional LSTM. In addition, neural network gate model to estimate the generation probability and the loss function to optimize the entire abstraction model has been applied. The dataset has been constructed from the collection of abstractions and titles of journal articles. Experimental results demonstrate that both ROUGE-1 (based on word recall) and ROUGE-L (employed longest common subsequence) of the proposed Encoding-Decoding model have been improved to 47.01 and 29.55, respectively.

Odd solutions to perturbed conservation laws

  • Jerome A. Goldstein;Park, Mi-Ai
    • Bulletin of the Korean Mathematical Society
    • /
    • v.33 no.4
    • /
    • pp.521-530
    • /
    • 1996
  • This paper treats the existence of odd solutions of the Cauchy problem for a perturbation of a conservation law.

  • PDF

Comparative Study of Make-up and Hair Styling Cultures of the King Jeongjo and the Edo Period (정조시대와 에도시대(江戶時代)의 화장문화(化粧文化)와 수발문화(鬚髮文化)의 비교 연구)

  • Kim, Min-Kyung
    • Korean Journal of Human Ecology
    • /
    • v.18 no.1
    • /
    • pp.189-200
    • /
    • 2009
  • In the Edo period speaking of hair culture, common women wore decorative pigtail ribbons on the right, and twisted from side to side their tressed hair in such a manner as ungeunmeori and traemeori. Instead of gachae, common women used gogae made of their own hair, ungeunmeori on forehead, or jjokjinmeori at the back of head. During the Edo period, people women naturally exposed their necklines as a way of exposing their faces in the aesthetically ceremonial act of wearing make-up. As for lipsticks, they rouged extracts from red petals of safflowers mainly on their lips, and sometimes on their cheeks by blending this with white powder. Samurai families disliked women who wore thick lip makeup. In the latter period, women painted their necklines or foreheads black, applied a small amount of rouge on their cheeks thinly or thickly, and colored a reddish color into their fingernails by using petals and leaves of balsam flowers. Despite the chronological and spatial proximity of the King Jeongjo period and the Edo period, it was found that there were no similarities between two countries' cosmetic cultures. Moreover, it was discovered that current TV dramas were being produced, even not based on historical evidence in the Jeongjo period.

Generation Paraphrase using Pointer Generation Network (포인터 생성 네트워크를 이용한 패러프레이즈 생성)

  • Park, Da-Sol;Kim, Young-kil;Cha, Jeong-Won
    • Annual Conference on Human and Language Technology
    • /
    • 2020.10a
    • /
    • pp.535-539
    • /
    • 2020
  • 다양한 발화를 모델링하는 요구는 자연어 처리 분야에서 꾸준히 있었으며 단어, 구 또는 문장과 동등한 의미 콘텐츠를 자동으로 식별하고 생성하는 것은 자연어 처리의 중요한 부분이다. 본 논문에서는 포인터 생성 네트워크(Pointer Generate Nework)를 이용하여 패러프레이즈 생성 모델을 제안한다. 제안한 모델의 성능을 측정하기 위해 사람이 직접 구축한 유사 문장 코퍼스를 이용하였으며, 토큰 단위의 BLEU-4 0.250, ROUGE_L 0.455, CIDEr 2.190의 성능을 보였다. 하지만 입력 문장과 동일한 문장을 출력하는 문제점이 존재하여 빔서치(beam search)를 적용하여 입력 문장과 비교하여 생성 문장을 선택하는 방식을 적용하였다. 입력 문장과 동일한 문장을 제외한 문장으로 평가를 진행했으며, 토큰 단위의 BLEU-4 0.234, ROUGE_L 0.459, CIDEr 2.041의 성능을 보였으나, 패러프레이즈 생성 데이터 양이 크게 증가하였다. 본 연구는 문장 간의 의미적으로 동일한 정보를 정확하게 추출할 수 있게 됨으로써 정보 추출, 온톨로지 생성에 도움이 될 것이다. 또한 이러한 기법이 챗봇에서 사용자의 의도 탐지 및 MRC와 같은 자연어 처리의 여러 분야에 유용한 자원으로 사용될 것이다.

  • PDF

Signifier of Father on the Traditional Fairy Tale『Le petit chaperon rouge『 and the Korean Film <Uncle> (전래동화 『빨간모자』와 영화 <아저씨>에 나타난 아버지의 기표)

  • Kim, Guyl-Hun
    • The Journal of the Korea Contents Association
    • /
    • v.12 no.7
    • /
    • pp.65-75
    • /
    • 2012
  • Throughout the agraian society, industrial society, information society, and the knowledge-based society, the word father has become the representation of the law, oder, control, protection, and discipline. This study, with traditional fairy tale "Le petit chaperon rouge", aim to analyze the child readers a notion familly around 'me' and conceiving the father as the unit of the institution; also, in the Korean film , cinema audiences feeling keenly the necessity of the conventional family values and looking for the poster of principal agent of law and order. Generally the father is protector of 'me' and simutaneously, is incorporal being; even if he is inessential, he has, only by the name, great influence with people(law, order, control, discipline, are operating only with the name as the father God does). Our study will show how the signifier father operates and what similarity is between the film and the traditional fairy tale. A father is recognized as a tyrant or criminal, but a man is always realized to undertake a duty of resisting unjust powers and protecting nonperson. Those opposing structures was well represented in epic works such as"Le petit chaperon rouge" and . The father described in children's tale appeared as a symbol of desire and oppression, therefore youngster readers realized around 'Me' the concept of family through the absence of a father and received the father as a unit in the system. A father in a film has been described as a protector of nonperson, and a main part in the institution and order of traditional family. In both genre, the psychological signifiant of father still has been shown based on the symbol of father as a social institution.