• ETRI Journal
• /
• v.39 no.4
• /
• pp.455-466
• /
• 2017

This study focuses on a method for sequential data augmentation in order to alleviate data sparseness problems. Specifically, we present corpus expansion techniques for enhancing the coverage of a language model. Recent recurrent neural network studies show that a seq2seq model can be applied for addressing language generation issues; it has the ability to generate new sentences from given input sentences. We present a method of corpus expansion using a sentence-chain based seq2seq model. For training the seq2seq model, sentence chains are used as triples. The first two sentences in a triple are used for the encoder of the seq2seq model, while the last sentence becomes a target sequence for the decoder. Using only internal resources, evaluation results show an improvement of approximately 7.6% relative perplexity over a baseline language model of Korean text. Additionally, from a comparison with a previous study, the sentence chain approach reduces the size of the training data by 38.4% while generating 1.4-times the number of n-grams with superior performance for English text.

• Journal of The Korean Association of Information Education
• /
• v.3 no.2
• /
• pp.113-119
• /
• 2000

I studied a distance education model for English learning on the Internet. Basic WWW files, that contain courseware, are constructed with HTML, and functions, which are required in learning, are implemented with Java. Students and educators can access the preferred unit composed of the appropriate text, voice and image data by using a WWW browser at any time. The education system supports the automatic generation facility of English problems to practice reading and writing by making good use of the courseware data or various English text resources located on the Internet. Our system has functions to manage and control the flow of distance learning and to offer interaction between students and the system in a distributed environment. Educators can manage students' learning and can immediately be aware of who is attending and who is quitting the lesson in virtual space. Also, students and educators in different places can communicate and discuss a topic through the server. I implemented these functions, which are required in a client/server environment of distance education, with the use of Java. The URL for this system is "http://park.taegu-e.ac.kr" in the name of GAIA.

• English Language & Literature Teaching
• /
• v.13 no.4
• /
• pp.153-165
• /
• 2007

The story treats each individual's vision as well as social reality that the author intends to describe. The purpose of this article is to search for the conflict between vision and reality, especially in feminist problem that critics have treated on the works of women writers. Though some articles have studied on the issue similar to this article, I try to analyze the narratives in the text that the author herself confesses to us. I think that we can find out clear messages from the individuals who construct the human relationship and build up their personal history through their dialogue or monologue. We can also catch their main problems in the community. I discuss the topic by mentioning the detailed discourses referred to the heroine and other characters in the text. The passages mentioned by the characters in the story may be a confession for the present and future generation that the author tries to confess. From the excerpts of some discourse, I can conclude that though Dorothea has a vision for her ideal, she is a failed feminist, for society is too strong for her as Miller (1990) argues.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1999.06a
• /
• pp.151-158
• /
• 1999

This paper describes TVML (TV program Making Language) for automatically generating television programs from text-based script. This language describes the contents of a television program using expression with a high level of abstraction like“title #1”and“zoom-in”. The software used to read a script written in TVML and to automatically generate the program video and audio is called the TVML Player. The paper begins by describing TVML language specifications and the TVML Player. It then describes the“external control mode”of the TVML Player that can be used for applying TVML to interactive applications. Finally, it describes the TVML Editor, a user interface that we developed which enables users having no specialized knowledge of computer languages to make TVML scripts. In addition to its role as a television-program production tool. TVML is expected to have a wide range of applications in the network and multimedia fields.

• Nuclear Engineering and Technology
• /
• v.54 no.6
• /
• pp.2062-2066
• /
• 2022

Nuclear energy sources are indispensable in cost effectively achieving carbon neutral economy, where public opinion is critical to adoption as the consequences of nuclear accident can be catastrophic. In this context, discussion on spent nuclear fuel is a prerequisite to expanding nuclear energy, as it leads to the issue of radioactive waste disposal. Given the dearth of study on spent nuclear fuel public acceptance, we use text mining and big data analysis on the news article and public comments data on Naver news portal to identify the Korean public opinion on spent nuclear fuel. We identify that the Korean public is more interested in the nuclear energy policy than spent nuclear fuel itself and that the alternative energy sources affect the position towards spent nuclear fuel. We recommend relating spent nuclear fuel issue with nuclear energy policy and environmental issues of alternative energy sources to further promote spent nuclear fuel.

• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.157-161
• /
• 2020

문서 요약(text summarization)은 주어진 문서로부터 중요하고 핵심적인 정보를 포함하는 요약문을 만들어 내는 작업으로, 기계 번역 작업에서 주로 사용되는 Sequence-to-Sequence 모델을 사용한 end-to-end 방식의 생성(abstractive) 요약 모델 연구가 활발히 진행되고 있다. 최근에는 BERT와 MASS 같은 대용량 단일 언어 데이터 기반 사전학습(pre-training) 모델을 이용하여 미세조정(fine-tuning)하는 전이 학습(transfer learning) 방법이 자연어 처리 분야에서 주로 연구되고 있다. 본 논문에서는 MASS 모델에 복사 메커니즘(copying mechanism) 방법을 적용하고, 한국어 언어 생성(language generation)을 위한 사전학습을 수행한 후, 이를 한국어 문서 요약에 적용하였다. 실험 결과, MASS 모델에 복사 메커니즘 방법을 적용한 한국어 문서 요약 모델이 기존 모델들보다 높은 성능을 보였다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2022.07a
• /
• pp.513-516
• /
• 2022

본 논문에서는 텍스트 마이닝 기법을 이용하여 MZ 세대를 이해하는 키워드를 도출하고자 한다. MZ 세대의 비중이 높아지면서, MZ 세대를 분석하려고 하는 많은 연구들이 수행되고 있다. 이에 본 연구에서는 MZ 세대를 이해하기 위하여 네이트 판의 연령별 게시판 크롤링을 통해 빅데이터를 수집하였다. 그리고 텍스트 마이닝 기법을 활용하여 10대, 20대, 30대의 각각의 키워드를 도출할 수 있었다. 본 논문에서 도출된 키워드는 이는 MZ 세대를 이해하는데 중요한 키워드로 볼 수 있을 것이다. 향후 연구로는 MZ 세대와 기성 세대를 비교하기 위하여 추가 크롤링을 통해 세대 간 비교 연구를 수행하고자 한다.

• International Journal of Computer Science & Network Security
• /
• v.24 no.2
• /
• pp.25-30
• /
• 2024

Various new technologies and aiding instruments are always being introduced for the betterment of the challenged. This project focuses on aiding the mute in expressing their views and ideas in a much efficient and effective manner thereby creating their own place in this world. The proposed system focuses on using various gestures traced into texts which could in turn be transformed into speech. The gesture identification and mapping is performed by the Kinect device, which is found to cost effective and reliable. A suitable text to speech convertor is used to translate the texts generated from Kinect into a speech. The proposed system though cannot be applied to man-to-man conversation owing to the hardware complexities, but could find itself very much of use under addressing environments such as auditoriums, classrooms, etc

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.574-575
• /
• 2024

본 논문은 T5(Text-To-Text Transfer Transformer) 모델을 활용한 반주 기반 가사 생성 기법을 제안하였다. 텍스트 이벤트 형식으로 변환한 정제된 반주를 "가사 생성" Task Token과 같이 T5에 적용하여 입력된 반주에 상응하는 가사를 생성하는 방식이다. 본 논문에서 제안한 방식의 성능 검증을 위해 Transformer, GPT-2, BART를 이용하여 가사를 생성한 출력물을 BLEU(Bilingual Evaluation Understudy) 값과 감정분석 일치도(Emotion Analysis Consistency) 결과값을 통해 비교 평가하였다. 본 논문에서 제안한 T5를 이용한 방식이 Transformer, GPT-2, BART를 사용하는 방식보다 우수한 결과를 얻었다.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.1
• /
• pp.19-26
• /
• 2023

In this study, the public sensitivity of the 'MZ generation' was examined through the social media big data sensitivity analysis method. For the analysis, the consumer account SNS text was examined, and positive and negative emotional factors were presented by classifying external sensibilities and emotions of the MZ generation. In conclusion, the positive emotions of liking and interest in relation to the "MZ generation" were 72.1%, higher than the negative emotional ratio of 27.9%. In positive sensitivity, the older generation showed 'a favorable feeling for the individuality and dignifiedness of the MZ generation' and 'interest in the MZ generation with new values'. In contrast, the MZ generation has a favorable feeling for 'the fact that they are a generation of their own boldness, youthfulness and individuality' and 'small growthism'. Negative sensitivity outside the MZ generation was found to be 'A concern about the marriage avoidance, employment difficulties, debt investment, and resignation trends of the MZ generation', 'Hate the MZ generation who treats Kkondae' and 'Difficult to talk to the MZ generation'. On the other hand, the negative emotions felt by the MZ generation itself were 'Rejection of generalization', 'Rejection of generation and gender conflicts', 'Rejection of competition worse than the older generation', 'Relative failure of the rich era', and 'Sadness to live in a predicted climate disaster'. Therefore, the older generation should not look at the MZ generation in general, but as individuals, and should alleviate conflicts with intergenerational understanding and empathy. there is a need for community consideration to solve generational conflicts, gender conflicts, and environmental problems.