• The Journal of Korean Association of Computer Education
• /
• v.12 no.2
• /
• pp.61-75
• /
• 2009

Modern day people can easily use an e-mail client for general communication, because of using Internet and cellular phone. The mail client for the purpose of private and business affair, advertisement, news searching, and business letter is widely used and has side effects. People could send an important document via an electronic mail client. It is important to support an e-mail client intelligent. We think that many kinds of techniques of natural language processing must be provided in the client with human's emotion. We consider to design a new mail client with six kinds of senders' emotional information; delight, angry, sad feeling and message to express, manner of talking, a discomfort index etc. Before sending an e-mail, we suggest a user to correct a bad word because we do not want to feel bad to a receiver. We present a proper process of sending/receiving for users with a new designed e-mail clients.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.11 no.8
• /
• pp.1604-1611
• /
• 2007

In a process of insect footprint recognition, basic footprint segments should be extracted from a whole insect footprint image in order to find out appropriate features for classification. In this paper, we used a clustering method as a preprocessing stage for extraction of basic insect footprint segments. In general, sizes and strides of footprints may be different according to type and sire of an insect for recognition. Therefore we proposed an improved ART2 algorithm for extraction or basic insect footprint segments regardless of size and stride or footprint pattern. In the proposed ART2 algorithm, threshold value for clustering is determined automatically using contour shape of the graph created by accumulating distances between all the spots of footprint pattern. In the experimental results applying the proposed method to two kinds of insect footprint patterns, we could see that all the clustering results were accomplished correctly.

• Journal of Internet Computing and Services
• /
• v.22 no.2
• /
• pp.69-75
• /
• 2021

In a contact-free society, online services are becoming more important than classic offline services. At the same time, the role of a contact center, which executes customer relation management (CRM), is increasingly essential. For supporting the CRM tasks and their effectiveness, techniques of process automation need to be applied. Quality assurance (QA) is one of the time and resource consuming, and typical processes that are suitable for automation. In this paper, a method of automatic quality evaluation for voice based consultations is proposed. Firstly, the speech in consultations is transformed into a text by speech recognition. Then quantitative evaluation based on the QA metrics, including checking the elements in opening and closing mention, the existence of asking the mandatory information, the attitude of listening and speaking, is executed. 92.7% of the automated evaluations are the same to the result done by human experts. It was found that the non matching cases of the automated evaluations were mainly caused from the mistranslated Speech-to-Text (STT) result. With the confidence of STT result, this proposed method can be employed for enhancing the efficiency of QA process in contact centers.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.16 no.2
• /
• pp.271-278
• /
• 2021

This paper is a study on the implementation of smart filter system that recognizes the external environment and automatically removes pollutants according to pollution level. Recently, the occurrence of various pollutants in indoor and outdoor space has adversely affected the human body. Especially, various fine dust generated in the atmosphere becomes worse in closed residential space or office space. Although air pollution can be temporary lowered through ventilation, it is difficult to respond to fine dust changes in real time, and such problems become serious in the space where many people reside, such as at home or industry. Therefore, it is necessary to measure the pollution level of fine dust inside the residential space in real time and to reduce the pollution of indoor ventilation through automatic ventilation with the outside. To improve these problems, this paper proposes the implementation of smart filter system for external environment recognition. The structure of smart filter system that automatically measures air quality inside and outside, removes pollutants, implements the function, and confirms the operability by manufacturing prototypes. Finally, the effectiveness of the smart filter system for solving fine dust problems was examined.

• The Journal of the Convergence on Culture Technology
• /
• v.9 no.1
• /
• pp.565-570
• /
• 2023

There are people who are sensitive to cosmetic ingredients, such as pregnant women and skin disease patients. There are also people who experience side effects from cosmetics. To avoid this, it is cumbersome to search for harmful ingredients in cosmetics one by one when shopping. In addition, knowing and remembering functional ingredients that suit you is helpful when purchasing new cosmetics. There is a need for a system that allows you to immediately know the cosmetics ingredients in the field through photography. In this paper, we introduce an application for smartphones, <Hwa Ahn>, which allows you to immediately know the cosmetics ingredients by photographing the ingredients displayed in the cosmetics. This system is more effective and convenient than the existing system in that it automatically recognizes and automatically classifies the ingredients of the cosmetic when the camera is illuminated on the cosmetic ingredients or retrieves the photos of the cosmetic ingredients from the album. If the system is widely used, it is expected that it will prevent skin diseases caused by cosmetics in daily life and reduce purchases of cosmetics that are not suitable for you.

• Annual Conference on Human and Language Technology
• /
• 2021.10a
• /
• pp.178-185
• /
• 2021

Sequence to sequence(S2S) 기반 음성인식 후처리기를 훈련하기 위한 학습 데이터 구축을 위해 (음성인식 결과(speech recognition sentence), 전사자(phonetic transcriptor)가 수정한 문장(Human post edit sentence))의 병렬 말뭉치가 필요하며 이를 위해 많은 노동력(human-labor)이 소요된다. BackTranScription (BTS)이란 기존 S2S기반 음성인식 후처리기의 한계점을 완화하기 위해 제안된 데이터 구축 방법론이며 Text-To-Speech(TTS)와 Speech-To-Text(STT) 기술을 결합하여 pseudo 병렬 말뭉치를 생성하는 기술을 의미한다. 해당 방법론은 전사자의 역할을 없애고 방대한 양의 학습 데이터를 자동으로 생성할 수 있기에 데이터 구축에 있어서 시간과 비용을 단축 할 수 있다. 본 논문은 BTS를 바탕으로 제주어 도메인에 특화된 음성인식 후처리기의 성능을 향상시키기 위하여 모델 수정(model modification)을 통해 성능을 향상시키는 모델 중심 접근(model-centric) 방법론과 모델 수정 없이 데이터의 양과 질을 고려하여 성능을 향상시키는 데이터 중심 접근(data-centric) 방법론에 대한 비교 분석을 진행하였다. 실험결과 모델 교정없이 데이터 중심 접근 방법론을 적용하는 것이 성능 향상에 더 도움이 됨을 알 수 있었으며 모델 중심 접근 방법론의 부정적 측면 (negative result)에 대해서 분석을 진행하였다.

• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.6
• /
• pp.29-43
• /
• 2023

Consumers should look at the ingredients of cosmetics or food for their health and purchase them after checking whether they contain allergy-causing ingredients. Therefore, this paper aimed to develop an artificial intelligence-based smartphone application for automatically recognizing the ingredients displayed on a curved container and delivering it to consumers in an easy-to-understand manner. The app needs to allow consumers to immediately comprehend the restricted ingredients by recognizing the ingredients' words in the cropped image. Two major issues should be solved during the development process: First, although there were flat containers for cosmetics or food, most were curved containers. Thus, it was necessary to recognize the ingredient table displayed on the curved containers. Second, since the ingredients' words were displayed on the curved surface, the transformed or line-changed words also needed to be recognized. The proposed new methods were enough to solve the above two problems. The application developed through various tests verified that there was no problem recognizing the ingredients' words contained in a cylindrical curved container.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.38C no.3
• /
• pp.278-287
• /
• 2013

Modern information technologies continue to provide industries with new and improved methods. With the rapid development of Machine to Machine (M2M) communication, a smart container supply chain management is formed based on high performance sensors, computer vision, Global Positioning System (GPS) satellites, and Globle System for Mobile (GSM) communication. Existing supply chain management has limitation to real time container tracking. This paper focuses on the studies and implementation of real time container chain management with the development of the container identification system and automatic alert system for interrupts and for normal periodical alerts. The concept and methods of smart container modeling are introduced together with the structure explained prior to the implementation of smart container tracking alert system. Firstly, the paper introduces the container code identification and recognition algorithm implemented in visual studio 2010 with Opencv (computer vision library) and Tesseract (OCR engine) for real time operation. Secondly it discusses the current automatic alert system provided for real time container tracking and the limitations of those systems. Finally the paper summarizes the challenges and the possibilities for the future work for real time container tracking solutions with the ubiquitous mobile and satellite network together with the high performance sensors and computer vision. All of those components combine to provide an excellent delivery of supply chain management with outstanding operation and security.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.4
• /
• pp.100-105
• /
• 2016

The use of 3D scanners has increased gradually according to increasing 3D printer applications. The precision inspection of car parts or electronic components is an important issue not only in the field of mass production, but also in small-scale production. Recently, 3D scanner equipment efficiency and recognition technology has been improved continuously. On the other hand, the spraying time to prepare 3D scanning is time-consuming and has environmental problems. Therefore, an automatic spray system has been in demand by the manufacturing industry. Automatic spray equipment was newly developed for the preparation of a 3D scanner. In this research, the automatic spray system guarantees uniform spray operation. To determine the optimal spray parameters, various spraying methods, solutions and conditions were tested and compared with the experiments. The preparation time for 3D scanning was reduced to 1/10 compared to the manual spraying time, and indicates the optimal spraying conditions through a comparison of various spray coating conditions.

• Journal of the Korean Society for information Management
• /
• v.40 no.2
• /
• pp.115-135
• /
• 2023

The purpose of this study is to assess the effectiveness of using deep learning language models to extract references automatically and create a reference database for research reports in an efficient manner. Unlike academic journals, research reports present difficulties in automatically extracting references due to variations in formatting across institutions. In this study, we addressed this issue by introducing the task of separating references from non-reference phrases, in addition to the commonly used metadata extraction task for reference extraction. The study employed datasets that included various types of references, such as those from research reports of a particular institution, academic journals, and a combination of academic journal references and non-reference texts. Two deep learning language models, namely RoBERTa+CRF and ChatGPT, were compared to evaluate their performance in automatic extraction. They were used to extract metadata, categorize data types, and separate original text. The research findings showed that the deep learning language models were highly effective, achieving maximum F1-scores of 95.41% for metadata extraction and 98.91% for categorization of data types and separation of the original text. These results provide valuable insights into the use of deep learning language models and different types of datasets for constructing reference databases for research reports including both reference and non-reference texts.