• 대한후두음성언어의학회지
• /
• 제32권3호
• /
• pp.109-117
• /
• 2021

The auditory-perceptual evaluation of speech-language pathologists (SLP) in patients with voice disorders is often regarded as a touchstone in the multi-dimensional voice evaluation procedures and provides important information not available in other assessment modalities. Therefore, it is necessary for the SLPs to conduct a comprehensive and in-depth evaluation of not only voice but also the overall speech production mechanism, and they often encounter various difficulties in the evaluation process. In addition, SLPs should strive to avoid bias during the evaluation process and to maintain a wide and constant spectrum of severity for each parameter of voice quality. Lastly, it is very important for the SLPs to perform a team approach by documenting and delivering important information pertaining to auditory-perceptual characteristics in an appropriate and efficient way through close communication with the laryngologists.

• International Journal of Computer Science & Network Security
• /
• 제22권7호
• /
• pp.352-358
• /
• 2022

Text mining is an important process used for analyzing the data collected from different sources like videos, audio, social media, and so on. The tools like Natural Language Processing (NLP) are mostly used in real-time applications. In the earlier research, text mining approaches were implemented using long-short memory (LSTM) networks. In this paper, text mining is performed using average-stochastic gradient descent weight-dropped (AWD)-LSTM techniques to obtain better accuracy and performance. The proposed model is effectively demonstrated by considering the internet movie database (IMDB) reviews. To implement the proposed model Python language was used due to easy adaptability and flexibility while dealing with massive data sets/databases. From the results, it is seen that the proposed LSTM plus weight dropped plus embedding model demonstrated an accuracy of 88.36% as compared to the previous models of AWD LSTM as 85.64. This result proved to be far better when compared with the results obtained by just LSTM model (with 85.16%) accuracy. Finally, the loss function proved to decrease from 0.341 to 0.299 using the proposed model

• Journal of Information Processing Systems
• /
• 제19권6호
• /
• pp.830-841
• /
• 2023

Tokenization is the process of segmenting the input text into smaller units of text, and it is a preprocessing task that is mainly performed to improve the efficiency of the machine learning process. Various tokenization methods have been proposed for application in the field of natural language processing, but studies have primarily focused on efficiently segmenting text. Few studies have been conducted on the Korean language to explore what tokenization methods are suitable for document classification task. In this paper, an exploratory study was performed to find the most suitable tokenization method to improve the performance of a representative spatio-temporal document classifier in Korean. For the experiment, a convolutional neural network model was used, and for the final performance comparison, tasks were selected for document classification where performance largely depends on the tokenization method. As a tokenization method for comparative experiments, commonly used Jamo, Character, and Word units were adopted. As a result of the experiment, it was confirmed that the tokenization of word units showed excellent performance in the case of representative spatio-temporal document classification task where the semantic embedding ability of the token itself is important.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권11호
• /
• pp.4184-4202
• /
• 2021

Text steganography combined with natural language generation has become increasingly popular. The existing methods usually embed secret information in the generated word by controlling the sampling in the process of text generation. A candidate pool will be constructed by greedy strategy, and only the words with high probability will be encoded, which damages the statistical law of the texts and seriously affects the security of steganography. In order to reduce the influence of the candidate pool on the statistical imperceptibility of steganography, we propose a steganography method based on a new sampling strategy. Instead of just consisting of words with high probability, we select words with relatively small difference from the actual sample of the language model to build a candidate pool, thus keeping consistency with the probability distribution of the language model. What's more, we encode the candidate words according to their probability similarity with the target word, which can further maintain the probability distribution. Experimental results show that the proposed method can outperform the state-of-the-art steganographic methods in terms of security performance.

• 한국정보관리학회:학술대회논문집
• /
• 한국정보관리학회 2006년도 제13회 학술대회 논문집
• /
• pp.163-168
• /
• 2006

Nursing diagnosis process is described as nurses assessing the patients' conditions by applying reasoning and looking for patterns, which fit the defining characteristics of one or more diagnoses. This process is similar to using a typical document retrieval system if we consider the patients' conditions as queries, nursing diagnoses as documents, and the defining characteristics as index terms of the documents. However, there is a small fixed number of nursing diagnoses and infinite number of patients' conditions in a typical hospital setting. This state is more suitable to applying document routing mode of information access, which is defined as a number of archived profiles, compared to individual documents. In this paper, we describe a ROUting-based Nursing Diagnosis (ROUND) system and its Natural Language Processing-based query processing component, which converts the defining characteristics of nursing diagnoses into query representations.

• 한국지능시스템학회논문지
• /
• 제11권3호
• /
• pp.246-251
• /
• 2001

본 연구에서는 성공적인 전자상거래시스템 관리를 위하여 지능적 검색 에이전트 모델을 제안했다. 퍼지 이론은 모호한 키워드 조건에서 시스템이 검색을 수행해야 할 경우에 매우 유용한 방법이다. 따라서, 퍼지 이론을 이용하여 고객의 모호한 검색어를 효과적으로 처리할 수 있는 모델을 제안했다. 이를 통해, 다른 크리스프한 검색어 환경에서의 시스템에 비해 상대적으로 적합한 결과를 도출할 수 있음을 확인했다.

• 한국컴퓨터정보학회논문지
• /
• 제23권1호
• /
• pp.17-24
• /
• 2018

With improvement of speech recognition and natural language processing, dialog systems are recently adapted to various service domains. It became possible to get desirable services by conversation through the dialog system, but it is still necessary to improve separate modules, such as domain detection, intention detection, named entity recognition, and out-of-domain detection, in order to achieve stable service offer. When it misclassifies an in-domain sentence of conversation as out-of-domain, it will result in poor customer satisfaction and finally lost business. As there have been relatively small number of studies related to the out-of-domain detection, in this paper, we introduce a new method using a hierarchical Dirichlet process and demonstrate the effectiveness of it by experimental results on Korean dataset.

• 한국정보통신학회:학술대회논문집
• /
• 한국정보통신학회 2022년도 추계학술대회
• /
• pp.72-74
• /
• 2022

국가과학기술지식정보서비스(이하 NTIS)에서는 국가R&D과제정보를 제공하고 있다. 과제정보는 '과제명', '과제수행기관', '연구책임자명' 등의 메타정보와 '연구목표', '연구내용', '기대효과'와 같은 과제를 설명하는 텍스트들로 구성되어있다. 과제정보 100만건을 대상으로 검색한 결과목록에서 '연구목표' 나 '연구내용' 등을 모두 확인하여 원하는 과제정보를 찾기 위해서는 많은 시간이 필요하다는 문제가 있다. 이러한 문제점을 해소하기 위해, 본 논문에서는 국가R&D 과제정보 내에서 장문의 텍스트로 구성된 부분을 요약하는 과제정보 요약 시스템을 제안하고자 한다. 한국어의 언어학적 특징을 분석하여 전처리기를 구축하고 전처리된 텍스트 정보를 처리하기 위한 자연어 처리 기술 기반 과제정보 요약 모델을 개발하였다. 이를 통해 장문으로 구성된 과제정보를 압축 및 요약된 형태로 제공하여, 이용자들이 요약정보만으로도 전반적인 내용을 쉽고 빠르게 유추하는 데 도움이 될 것이다.

• 한국컴퓨터정보학회논문지
• /
• 제28권11호
• /
• pp.81-87
• /
• 2023

네트워크 패킷의 메타데이터를 학습한 침입탐지시스템이 최근 많이 제안되었다. 그러나 이러한 방식은 모델 학습에 사용할 메타데이터 생성을 위해 패킷을 분석하는 시간, 그리고 학습 전 메타데이터를 전처리하는 시간이 필요하다. 또한, 특정 메타데이터를 학습한 모델은 실제 네트워크로 유입되는 원본 패킷을 그대로 사용하여 침입을 탐지하는 것이 불가능하다. 이러한 문제를 해결하기 위해 본 논문에서는 패킷 페이로드를 하나의 문장으로 학습시켜 침입을 탐지하는 자연어 처리 기반의 침입탐지시스템을 제안하였다. 제안하는 기법의 성능 검증을 위해 UNSW-NB15와 Transformer 모델을 활용하였다. 먼저, 데이터세트의 PCAP 파일에 대한 라벨링을 실시한 후 2종의 Transformer 모델(BERT, DistilBERT)에 문장 형태로 직접 학습시켜 탐지성능을 분석하였다. 실험 결과 이진분류 정확도는 각각 99.03%, 99.05%로 기존 연구에서 제안한 기법들과 유사하거나 우수한 탐지성능을 보였으며, 다중분류는 각각 86.63%, 86.36%로 더 우수한 성능을 나타냄을 확인하였다.

• 디지털융복합연구
• /
• 제18권11호
• /
• pp.215-226
• /
• 2020

챗봇이 비즈니스의 중요한 인터페이스 창구로 떠오르고 있다. 이러한 변화는 챗봇 관련 연구가 자연어처리(Natural Language Processing)기법에서 자연어이해(Natural Language Understanding) 그리고 자연어생성(Natural Language Generation)으로 지속적으로 발전했기 때문이다. 하지만, 챗봇을 개발하는 과정에서 도메인 지식을 이끌어내고, 사용자 친화적인 대화형 인터페이스로 개발하는 방법론적 연구는 미약한 것이 현실이다. 본 논문에서는 챗봇 개발의 프로세스적 기준을 제시하기 위해 이전 논문에서 제시한 방법론을 바탕으로 실제 프로젝트에 적용하며 개발방법론을 개선하였다. 결론적으로 가장 핵심적인 단계인 테스트 단계의 생산성을 33.3% 향상하였으며, 그 반복횟수도 37.5%로 단축하였다. 이러한 결과를 바탕으로 "3 Phase and 17 Tasks 개발방법론"을 제시하였으며, 이것은 챗봇 개발의 시행착오를 획기적으로 개선할 것으로 기대한다.