• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.267-268
• /
• 2022

Recently, big data analysis can use various techniques according to the development of machine learning. Big data collected in reality lacks an automated refining technique for the same or similar terms based on semantic analysis of the relationship between words. Big data is usually in the form of sentences, and morphological analysis or understanding of the sentences is required. Accordingly, NLP, a technique for analyzing natural language, can understand the relationship of words and sentences. In this paper, we study the advantages and disadvantages of Transformers and Reformers, which are techniques that complement the disadvantages of RNN, which is a time series approach to big data.

• Annual Conference on Human and Language Technology
• /
• 1995.10a
• /
• pp.300-309
• /
• 1995

언어표현에 대한 해석은 그 구성요소들의 통사적-어휘적 구조에 덧붙여 대명사의 동일지시를 살펴야 할 필요가 있다. 조응의 분석과 조응적 선행사를 찾기 위한 효과적인 방법을 발견하는 것이 컴퓨터 언어학(computational linguistics), 특히 자연언어 이해체계(Natural Language understanding system)에 관한 연구의 중심적인 문제라고 할 수 있다. 이 논문의 목적은 영어 조응표현을 개념구조 이론(Conceptual Structure Theory)의 개념도식(conceptual graph)에 의하여 기술함으로써 단문에서뿐만 아니라 복문, 양화구문, 그리고 담화에 이르기까지 언어 전반에 걸쳐 나타나는 동일지시성(coreferenciality)을 간단하고 일관성 있게 설명하는 것이다. 이러한 조응현상을 설명하기 위하여 필자는 개념도식상의 개념을 중심개념, 직접개념, 간접개념으로 구분하고 이들이 문맥깊이 등과 더불어 동일지시성을 설명하는데 중심적 역할을 함을 보이고자 한다.

• Annual Conference on Human and Language Technology
• /
• 1992.10a
• /
• pp.637-646
• /
• 1992

컴퓨터를 이용하여 자연언어를 처리하려는 연구가 진행되고 있다. 언어가 사고와 밀접한 관계에 있다는 점에서 이러한 연구가 성공한다면 인공지능의 발전과 더불어 인간의 마음에 대한 이해의 폭을 넓히게 될 것이다. 이글은 이러한 연구와 관련하여 컴퓨터에 의한 자연언어 이해의 가능성을 다루고 있다. 먼저 그러한 이해가 불가능하다는 써얼의 비판을 시발로 해서, 써얼에 대한 라파포트의 재반박을 검토할 것이다. 라파포트는 자신의 인공지능 프로그램과 사고실험 등을 통해서 가능성을 인정한다. 그의 주장의 핵심은 컴퓨터가 자연언어를 이해하는데 있어 구문론적 이해만으로도 충분하다는 것이다. 이러한 주장은 기호학적 관점에서 볼 때 성립될 수 없다고 비판된다. 인간이나 컴퓨터가 자연언어를 이해하기 위해서는 언어와 그 지시 대상, 그리고 언어의 사용자간의 관계를 고려하는 기호학적 관점이 요구된다. 그결과 컴퓨터에 의한 자연언어 이해의 가능성에는 한계가 있다는 결론에 이르게 된다.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.165-169
• /
• 2019

자연어 이해 모델은 대화 시스템의 핵심적인 구성 요소로서 자연어 문장에 대해 그 의도와 정보를 파악하여 의도(intent)와 슬롯(slot)의 형태로 분석하는 모델이다. 최근 연구에서 의도와 슬롯의 추정을 단일 합동 모델(joint model)을 이용하여 합동 학습(joint training)을 하는 연구들이 진행되고 있다. 합동 모델을 이용한 합동 학습은 의도와 슬롯의 추정 정보가 모델 내에서 암시적으로 교류 되도록 하여 의도와 슬롯 추정 성능이 향상된다. 본 논문에서는 기존 합동 모델이 암시적으로 추정 정보를 교류하는 데서 더 나아가 모델 내의 의도와 슬롯 추정 정보를 명시적으로 교류하도록 모델링하여 의도와 슬롯 추정 성능을 높일 수 있는 교차 게이트 메커니즘(Cross Gated Mechanism)을 제안한다.

• The Korean Journal of Elementary Counseling
• /
• v.7 no.1
• /
• pp.205-226
• /
• 2008

The purpose of the study you have seen is to verify the effectiveness of existing quantitative research and to put the Empathy Training Program to practical use for participating children. From looking into this, the changes in empathic understanding that came to light in relationships between teacher and children and children and children are sure to have that effect. For this work, I established the following subject of inquiry: What kind of changing processes can be seen in the empathic understanding of participating children in the Empathy Training Program? To resolve the above line of inquiry, six female sixth grade elementary school students were chosen and they progressed through twelve sessions of the Empathy Training Program. The children were given a sentence completion exam, recognition work, neat writing exam and a school adaptation exam both before and after participation in the program, making data for analysis. To analyze, first, participants had one or two meetings of forty to fifty minutes each. Progress through the program's curriculum was recorded and through the repeating and copying method, to be sure participating children's empathic understanding was revealed, empathic language and behavior was routinely chosen. Next, according the above criteria I looked into visible changes of the participating children's empathic expressions, classifying and analyzing changes in empathic understanding and six instances of common changes in the emphatic understanding of the participants relationships were analyzed and put together. Next I will summarize the findings we have seen in this research: First, if we look into changes in common empathic understanding from the beginning, using the criteria of empathic language, each individual showed understanding at the beginning and passed and progressed through stages of care, insight and emotional expressions. Second, when we looked at the criteria of empathic behavior from the beginning to the end, one's line of vision and ability to concentrate one's attention was connected. Next, the act of nodding one's head looked like a brief nod at first but at the end, it was not just a simple nod but rather they could feel deep empathy. The condition and substance of the facial expression was seen to match and at the very end the child was expressive and stretched out arms to hold and pat the other person and the act of holding hands could also be seen. Among lots of empathic behavior the final stage was shown by half of the children. Third, from the first stage to the last stage there were many cases revealed. The more the children went the more complete their empathic language became. Their vocabulary increased and became more diverse with empathic actions. Also, when comparing actions and expressions from the beginning with the end, visible expressions became more natural and sincere at the end. The result of the research we have seen is that through receiving experience of empathic understanding, participating children showed a sense of self-confidence and they looked to make peaceful expressions while not being aggressive or defensive about problems. In addition, from understanding empathic expressions, participating children's relationships felt closer. This outcome within this group in this case will be applied and the formation of empathic understanding can be used by the children internally to solve their own problems, acquire close relationships with their teachers and others. It will also contribute to smooth classroom management.

• Journal of the Korea Convergence Society
• /
• v.13 no.1
• /
• pp.13-21
• /
• 2022

Recent deep learning-based research shows excellent performance in most natural language processing (NLP) fields with pre-trained language models. In particular, the auto-encoder-based language model proves its excellent performance and usefulness in various fields of Korean language understanding. However, the decoder-based Korean generative model even suffers from generating simple sentences. Also, there is few detailed research and data for the field of conversation where generative models are most commonly utilized. Therefore, this paper constructs multi-turn dialogue data for a Korean generative model. In addition, we compare and analyze the performance by improving the dialogue ability of the generative model through transfer learning. In addition, we propose a method of supplementing the insufficient dialogue generation ability of the model by extracting recommended response candidates from external knowledge information through a retrival model.

• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.5
• /
• pp.31-39
• /
• 2023

The evolution of Natural Language Processing (NLP) and the rise of large language models (LLM) like ChatGPT have paved the way for specialized question-answering (QA) systems tailored to specific domains. This study outlines a system harnessing the power of LLM in conjunction with document search algorithms to interpret and address user inquiries using documents from the Korea Atomic Energy Research Institute (KAERI). Initially, the system refines multiple documents for optimized search and analysis, breaking the content into managable paragraphs suitable for the language model's processing. Each paragraph's content is converted into a vector via an embedding model and archived in a database. Upon receiving a user query, the system matches the extracted vectors from the question with the stored vectors, pinpointing the most pertinent content. The chosen paragraphs, combined with the user's query, are then processed by the language generation model to formulate a response. Tests encompassing a spectrum of questions verified the system's proficiency in discerning question intent, understanding diverse documents, and delivering rapid and precise answers.

• Korean Journal of Cognitive Science
• /
• v.2 no.2
• /
• pp.309-339
• /
• 1990

The most common way that people communicate is by speaking or writing natural languages.But if people use computers in the modern technology,they should learn artificial programming languages.If computers could understand what people mean when people speak or type natural languages,people would use the computers more easily and naturally.but there is a problem.The language which people use has vagueness.For example,the convential computer system cant's handle the subjective feeling like 'tall' or 'young'.So peole must specify the exact threshold like 'more'than 25 ages'.We have developed the knowledge-based natural language question answering system which can handle sentences having fuzzy concepts by using blackboard model.Our goal of this research is to develop a portable question answering system as interface for database systems or understanding systems.

• The Journal of Bigdata
• /
• v.6 no.1
• /
• pp.63-81
• /
• 2021

The performance of natural language processing is rapidly improving due to the recent development and application of machine learning and deep learning technologies, and as a result, the field of application is expanding. In particular, as the demand for analysis on unstructured text data increases, interest in NLP(Natural Language Processing) is also increasing. However, due to the complexity and difficulty of the natural language preprocessing process and machine learning and deep learning theories, there are still high barriers to the use of natural language processing. In this paper, for an overall understanding of NLP, by examining the main fields of NLP that are currently being actively researched and the current state of major technologies centered on machine learning and deep learning, We want to provide a foundation to understand and utilize NLP more easily. Therefore, we investigated the change of NLP in AI(artificial intelligence) through the changes of the taxonomy of AI technology. The main areas of NLP which consists of language model, text classification, text generation, document summarization, question answering and machine translation were explained with state of the art deep learning models. In addition, major deep learning models utilized in NLP were explained, and data sets and evaluation measures for performance evaluation were summarized. We hope researchers who want to utilize NLP for various purposes in their field be able to understand the overall technical status and the main technologies of NLP through this paper.

• ETRI Journal
• /
• v.43 no.2
• /
• pp.299-312
• /
• 2021

It is necessary to achieve high performance in the task of zero anaphora resolution (ZAR) for completely understanding the texts in Korean, Japanese, Chinese, and various other languages. Deep-learning-based models are being employed for building ZAR systems, owing to the success of deep learning in the recent years. However, the objective of building a high-quality ZAR system is far from being achieved even using these models. To enhance the current ZAR techniques, we fine-tuned a pretrained bidirectional encoder representations from transformers (BERT). Notably, BERT is a general language representation model that enables systems to utilize deep bidirectional contextual information in a natural language text. It extensively exploits the attention mechanism based upon the sequence-transduction model Transformer. In our model, classification is simultaneously performed for all the words in the input word sequence to decide whether each word can be an antecedent. We seek end-to-end learning by disallowing any use of hand-crafted or dependency-parsing features. Experimental results show that compared with other models, our approach can significantly improve the performance of ZAR.