• 제어로봇시스템학회논문지
• /
• 제9권7호
• /
• pp.498-506
• /
• 2003

A computer vision system applied to an intelligent safety vehicle has been required to be worked on a small sized real time special purposed hardware not on a general purposed computer. In addition, the system should have a high reliability even under the adverse road traffic environment. This paper presents a design and an implementation of an onboard hardware system taking into account for high speed image processing to analyze a road traffic scene. The system is mainly composed of two parts: an early processing module of FPGA and a postprocessing module of DSP. The early processing module is designed to extract several image primitives such as the intensity of a gray level image and edge attributes in a real-time Especially, the module is optimized for the Sobel edge operation. The postprocessing module of DSP utilizes the image features from the early processing module for making image understanding or image analysis of a road traffic scene. The performance of the proposed system is evaluated by an experiment of a lane-related information extraction. The experiment shows the successful results of image processing speed of twenty-five frames of 320$\times$240 pixels per second.

• 한국정보처리학회논문지
• /
• 제3권3호
• /
• pp.417-428
• /
• 1996

본 논문에서는 자연언어 대화를 통한 데이타베이스 인터페이스 시스템을 설계 및 구현한다. 시스템은 크게 언어해석부, 문맥처리부, 대화처리부,및 데이타베이스 처리 로 구성된다. 언어해석에 있어 입력 발화문의 미정의어를 분류하여 처리하는 방법을 제시함으로써 데이타 베이스 인터페이스의 난점으로 지적되어 온 사전의 크기를 효과 적으로 줄일 수 있었다. 그리고 기존의 개별적인 임력 발화문에 의한 자연언어 검색 시스템과는 달리 본 시스템은 대화 처리부를 통하여 한 문장의 입력문이 아닌 연속된 대화를 통하여 데이타베이스 정보를 검색할 수 있는 인터페이스 환경을 제공한다. 라서 본 논문에서는 명제적 내용뿐 아니라 사용자의 의도가 포함된 화행을 정의하고 따문헌 검색을 위한 사용자 행위 모델을 설정하였다. 따라서 시스템은 이와 같은 지식들 을 이용하여 사용자의 계획을 인식하였으며 효율적으로 대화 이해 및관리를 수행할 수 있었다.그리고 본논문에서제안된 방법론의 검증을 위하여 문헌 데이타베이스 영역에서 시스템을 구현하였다.

• 근관절건강학회지
• /
• 제28권3호
• /
• pp.252-262
• /
• 2021

Purpose: To define the concept of digital health literacy and identify its attributes. Methods: Walker and Avant's approach was employed for concept analysis. Attributes, antecedents, consequences, and the definition of digital health literacy were derived from a review of 28 studies. Results: Digital health literacy was identified to possess the following five attributes: health information seeking, health information processing, health information communication, health-related knowledge translation, and utilizing digital technology. Basic literacy skills, health concerns, motivation to use technology for health information, and access to digital technologies were all antecedents of the concept. The consequences of the concept were health behaviors, patient engagement, health status, and quality of life. Digital health literacy is the ability to seek relevant health information utilizing digital technology to solve health problems and improve quality of life. Furthermore, it refers to the translation of health-related knowledge obtained through health information processing-finding, understanding, and evaluating health information and health information communication-into the context in which individual and social factors interact. Conclusion: This study presented a new definition of digital health literacy that goes beyond existing internet-based eHealth literacy, by incorporating the context of emerging digital technologies. This proposed definition can serve as a foundation for the development of instruments and educational programs to improve individuals' digital health literacy.

• Journal of Information Processing Systems
• /
• 제12권2호
• /
• pp.234-248
• /
• 2016

Due to the proliferation of data being exchanged and the increase of dependency on this data for critical decision-making, it has become imperative to ensure the trustworthiness of the data at the receiving end in order to obtain reliable results. Data provenance, the derivation history of data, is a useful tool for evaluating the trustworthiness of data. Various frameworks have been proposed to evaluate the trustworthiness of data based on data provenance. In this paper, we briefly review a history of these frameworks for evaluating the trustworthiness of data and present an overview of some prominent state-of-the-art evaluation frameworks. Moreover, we provide a comparative analysis of two key frameworks by evaluating various aspects in an executional environment. Our analysis points to various open research issues and provides an understanding of the functionalities of the frameworks that are used to evaluate the trustworthiness of data.

• Journal of Information Processing Systems
• /
• 제19권6호
• /
• pp.803-816
• /
• 2023

Explainable artificial intelligence is a method that explains how a complex model (e.g., a deep neural network) yields its output from a given input. Recently, graph-type data have been widely used in various fields, and diverse graph neural networks (GNNs) have been developed for graph-type data. However, methods to explain the behavior of GNNs have not been studied much, and only a limited understanding of GNNs is currently available. Therefore, in this paper, we propose an explanation method for node classification using graph convolutional networks (GCNs), which is a representative type of GNN. The proposed method finds out which features of each node have the greatest influence on the classification of that node using GCN. The proposed method identifies influential features by backtracking the layers of the GCN from the output layer to the input layer using the gradients. The experimental results on both synthetic and real datasets demonstrate that the proposed explanation method accurately identifies the features of each node that have the greatest influence on its classification.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2005년도 춘계학술발표대회
• /
• pp.651-654
• /
• 2005

Ontology developing process has aroused a lot of controversy among knowledge engineers and knowledge users. The recent surges on ontology building methodologies and practical ontology applications have explored a broad spectrum of knowledge management challenges. On the one hand, the abundant methodology theories provide us with a set of useful heuristic rules, from which we get the overview of ontology building process. But on the other hand, every research groups would like to justify their theories by listing their specific characteristics and unique method when approaching the right way. However, there is still no one “correct” way or methodology for developing ontologies. In this case, the methods used to evaluate only a subset of specific domain do not make any sense to the commonsense users. As a result, a comprehensive understanding of domain ontology is urgent and necessary.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2009년도 추계학술발표대회
• /
• pp.803-804
• /
• 2009

To accommodate constantly growing knowledge in scientific discourse that is revised over time by domain experts, we need to also evolve our ontology. The body of knowledge will get structured and refined as we develop a deeper understanding of issues. Keeping trail of new changes in semantically rich and formally sound mechanism has pragmatic advantages for providing the undo and redo facility and ontology recovery to a previous state. In this research, we have proposed a framework that support change logging and then using these logged changes for reverting ontology to a previous consistent state and visualization of change effects on ontology. The system is compared with ChangesTab of $Prot{\acute{e}}g{\acute{e}}$ and the results depict better accuracy for our system.

• 한국정보처리학회논문지
• /
• 제7권5호
• /
• pp.1456-1463
• /
• 2000

The main objective of this study is to show how to measure the education effectiveness of simulator system in the army. We investigate the education effectiveness through the measuring model that consists of three dimensions such as understanding, experience, and learning. The results of empirical analysis demonstrate that the education effectiveness depends on three dimensions. The result also suggests that the higher the degree of each dimension is, the higher the education effectiveness. However, there is no education effectiveness difference between the traditional training and the simulator training because of the elementary level of simulator function.

• Journal of Information Processing Systems
• /
• 제11권4호
• /
• pp.528-537
• /
• 2015

Text in images is one of the most important cues for understanding a scene. In this paper, we propose a novel approach based on interest points to localize text in natural scene images. The main ideas of this approach are as follows: first we used interest point detection techniques, which extract the corner points of characters and center points of edge connected components, to select candidate regions. Second, these candidate regions were verified by using tensor voting, which is capable of extracting perceptual structures from noisy data. Finally, area, orientation, and aspect ratio were used to filter out non-text regions. The proposed method was tested on the ICDAR 2003 dataset and images of wine labels. The experiment results show the validity of this approach.

• 한국정보처리학회:학술대회논문집
• /
• 한국정보처리학회 2012년도 추계학술발표대회
• /
• pp.339-342
• /
• 2012

Nowadays, Spoken Dialogue System is rapidly growing by investing a lot from researches as well as organizations. One of concrete evidences is that the appearance of commercial systems such as Siri, SVoice, DARPA, CLASSiC, GSearch etc. Moreover, Spoken Dialogue System is widely believed to be the future direction of software development. In Spoken Dialogue System, users interact to software by using their own voice instead of use their hands, keyboard, and mouse. This paper continuously presents our development of the Spoken Dialogue System on News. Particularly, we propose detailed design such as semantic concepts, semantic frames, slots, and so on for applying Hidden Vector State Model into our Spoken Dialogue System for Spoken Language Understanding.