• Annual Conference on Human and Language Technology
• /
• 2013.10a
• /
• pp.9-12
• /
• 2013

본 논문에서는 한국어 형태소 분할 및 품사 태깅 방법을 위한 결합 모델로 Semi-CRF와 Linear-chain CRF에 대한 초기 비교 실험을 수행한다. Linear-chain방법은 출력 레이블을 형태소 분할 정보와 품사 태그를 조합함으로써 결합을 시도하는 방식이고, Semi-CRF는 출력의 구조가 분할과 태깅 정보를 동시에 포함하도록 표현함으로써, 디코딩 과정에서 분할과 태깅을 동시에 수행하는 방법이다. Sejong품사 부착말뭉치에서 비교결과 Linear-chain방법이 Semi-CRF방법보다 우수한 성능을 보여주었다.

• International Journal of Computer Science & Network Security
• /
• v.22 no.8
• /
• pp.55-60
• /
• 2022

Specific Language Impairment is one of the serious disorders that interferes with spontaneous communication skills in children. Children suffering from this disorder may have reading, speaking, or listening impairments, and such type of disorders are also termed Autism Speech Disorder (ASD) in medical terminology. The aim of the article is to define specific language impairment in children and the problems it can cause. The different methods adopted by speech pathologists to diagnose language impairment. Finally implementing machine learning models to automate the process and help speech pathologists and pediatricians/ in diagnosing the specific language impairment.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.433-438
• /
• 2021

대화 시스템은 크게 사용자와 시스템이 특정 목적 혹은 자유 주제에 대해 대화를 진행하는 것으로 구분된다. 최근 자유주제 대화 시스템(Open-Domain Dialogue System)에 대한 연구가 활발히 진행됨에 따라 자유 주제를 기반으로 하는 상담 대화, 일상 대화 시스템의 독성 발화 제어 생성에 대한 연구의 중요성이 더욱 커지고 있다. 이에 본 논문에서는 대화 모델의 독성 응답 생성을 제어하기 위해 일상 대화 데이터셋으로 학습된 BART 모델에 Plug-and-Play Language Model 방법을 적용한다. 공개된 독성 대화 분류 데이터셋으로 학습된 독성 응답 분류기를 PPLM의 어트리뷰트(Attribute) 모델로 활용하여 대화 모델의 독성 응답 생성을 감소시키고 그 차이를 실험을 통해 정량적으로 비교한다. 실험 결과 어트리뷰트 모델을 활용한 모든 실험에서 독성 응답 생성이 감소함을 확인하였다.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.66-71
• /
• 2023

ChatGPT가 등장한 이후 다양한 거대 언어 모델(Large Language Model, LLM)이 등장하였고, 이러한 LLM을 목적에 맞게 파인튜닝하여 사용할 수 있게 되었다. 하지만 LLM을 새로 학습하는 것은 물론이고, 단순 튜닝만 하더라도 일반인은 시도하기 어려울 정도의 많은 컴퓨팅 자원이 필요하다. 본 연구에서는 공개된 LLM을 별도의 학습 없이 사용하여 zero-shot 프롬프팅으로 이진 분류 태스크에 대한 성능을 확인하고자 했다. 학습이나 추가적인 튜닝 없이도 기존 선학습 언어 모델들에 준하는 이진 분류 성능을 확인할 수 있었고, 성능이 좋은 LLM의 경우 분류 실패율이 낮고 일관적인 성능을 보여 상당히 높은 활용성을 확인하였다.

• Journal of Convergence Society for SMB
• /
• v.4 no.4
• /
• pp.31-36
• /
• 2014

Recent advancement in Wireless Sensor Networks (WSNs) has paved the way for WSNs to enable in various environments in monitoring temperature, motion, sound, and vibration. These applications often include the detection of sensitive information from enemy movements in hostile areas or in locations of personnel in buildings. Due to characteristics of WSNs and dealing with sensitive information, wireless sensor nodes tend to be exposed to the enemy or in a hazard area, and security is a major concern in WSNs. Because WSNs pose unique challenges, traditional security techniques used in conventional networks cannot be applied directly, many researchers have developed various security protocols to fit into WSNs. To develop countermeasures of various attacks in WSNs, descriptions and analysis of current security attacks in the network layers must be developed by using a standard notation. However, there is no research paper describing and analyzing security models in WSNs by using a standard notation such as The Unified Modeling Language (UML). Using the UML helps security developers to understand security attacks and design secure WSNs. In this research, we provide standard models for security attacks by UML Sequence Diagrams to describe and analyze possible attacks in the three network layers.

• Phonetics and Speech Sciences
• /
• v.15 no.3
• /
• pp.83-88
• /
• 2023

This study investigates the fine-tuning of large-scale Automatic Speech Recognition (ASR) models, specifically OpenAI's Whisper model, for domain-specific applications using the KsponSpeech dataset. The primary research questions address the effectiveness of targeted lexical item emphasis during fine-tuning, its impact on domain-specific performance, and whether the fine-tuned model can maintain generalization capabilities across different languages and environments. Experiments were conducted using two fine-tuning datasets: Set A, a small subset emphasizing specific lexical items, and Set B, consisting of the entire KsponSpeech dataset. Results showed that fine-tuning with targeted lexical items increased recognition accuracy and improved domain-specific performance, with generalization capabilities maintained when fine-tuned with a smaller dataset. For noisier environments, a trade-off between specificity and generalization capabilities was observed. This study highlights the potential of fine-tuning using minimal domain-specific data to achieve satisfactory results, emphasizing the importance of balancing specialization and generalization for ASR models. Future research could explore different fine-tuning strategies and novel technologies such as prompting to further enhance large-scale ASR models' domain-specific performance.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.5
• /
• pp.109-116
• /
• 2022

Recently, a model-based development method centered on highly readable and standardized UML (Unified Modeling Language) models has been applied to solve unclear communications in large-scale software development. However, it is difficult to apply consistent UML models depending on software developers' proficiency, understanding of models and modeling tools. In this paper, we propose a method for developing a Model Verification System to apply an consistent UML model to software development. Then, the developed Model Verification System is partially applied to the Naval Combat System Software development to prove its function. The Model Verification System provides automatic verification of models created by developers according to domain characteristics. If the Model Verification System proposed in this paper is used, It has the advantage of being able to apply the consistent UML model more easily to Naval Combat System Software Development.

• Journal of English Language & Literature
• /
• v.55 no.3
• /
• pp.497-515
• /
• 2009

Culture as a source of identity, as Edward Said says, can be a battleground on which various political and ideological causes engage one another. It is not mere individual cultivation or private possession, but a program for social cohesion. Sensitively aware that a national culture should be independent from Europe, Walt Whitman enacts a new form of literature by placing different cultural values against Old World tradition. His interest in autochthonous culture originates from his deep concern about national consciousness. He believes that literary taste directed toward highly-ornamented elite culture is an obstacle to cultural unification of a nation. In order to represent American culture of the common people, Whitman incorporates a lot of cultural material into his poetry. Since he believes that America has many respectable writers at home, he urges people to adjust to their own taste instead of running after foreign authors. Whitman differentiated his poetry from previous literary models by disrupting the established literary norms and reconfiguring cultural values on the basis of American ways of life. In his comment on other poets, he concentrates on the originality and nativity of poetry. By claiming that words have characteristics of nativity, independence, and individuality, he envisions American literature to be distinguished from British literature in literary materials as well as in language. Whitman s language is composed of a vast number of words that can fully portray the nation. He works over language materials in two ways: reterritorialization and deterritorialization. Not only does his literary language become subversive of the established literary language, but also makes it possible to express strength and intensity in feeling.

• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.3
• /
• pp.16-23
• /
• 2001

In this paper, we propose the pseudo n-gram language models for speech recognition with middle size vocabulary compared to large vocabulary speech recognition using the statistical n-gram language models. The proposed method is that it is very simple method, which has the standard structure of ARPA and set the word probability arbitrary. The first, the 1-gram sets the word occurrence probability 1 (log likelihood is 0.0). The second, the 2-gram also sets the word occurrence probability 1, which can only connect the word start symbol and WORD, WORD and the word end symbol . Finally, the 3-gram also sets the ward occurrence probability 1, which can only connect the word start symbol , WORD and the word end symbol . To verify the effectiveness of the proposed method, the word recognition experiments are carried out. The preliminary experimental results (off-line) show that the word accuracy has average 97.7% for 452 words uttered by 3 male speakers. The on-line word recognition results show that the word accuracy has average 92.5% for 20 words uttered by 20 male speakers about stock name of 1,500 words. Through experiments, we have verified the effectiveness of the pseudo n-gram language modes for speech recognition.

• Korean Journal of Cognitive Science
• /
• v.35 no.1
• /
• pp.23-48
• /
• 2024

Building human-aligned artificial intelligence (AI) for social support remains challenging despite the advancement of Large Language Models. We present a novel method, the Chain of Empathy (CoE) prompting, that utilizes insights from psychotherapy to induce LLMs to reason about human emotional states. This method is inspired by various psychotherapy approaches-Cognitive-Behavioral Therapy (CBT), Dialectical Behavior Therapy (DBT), Person-Centered Therapy (PCT), and Reality Therapy (RT)-each leading to different patterns of interpreting clients' mental states. LLMs without CoE reasoning generated predominantly exploratory responses. However, when LLMs used CoE reasoning, we found a more comprehensive range of empathic responses aligned with each psychotherapy model's different reasoning patterns. For empathic expression classification, the CBT-based CoE resulted in the most balanced classification of empathic expression labels and the text generation of empathic responses. However, regarding emotion reasoning, other approaches like DBT and PCT showed higher performance in emotion reaction classification. We further conducted qualitative analysis and alignment scoring of each prompt-generated output. The findings underscore the importance of understanding the emotional context and how it affects human-AI communication. Our research contributes to understanding how psychotherapy models can be incorporated into LLMs, facilitating the development of context-aware, safe, and empathically responsive AI.