• Journal of Digital Convergence
• /
• v.10 no.11
• /
• pp.427-432
• /
• 2012

The article herein suggested a compound repository and a descriptive method to develop a compound knowledge process. A data target saved in a compound knowledge repository suggested in this article includes all compound knowledge meta data and digital resources, which can be divided into the three following factors according to the purpose: user roles, functional elements, and service ranges. The three factors are basic components to describe abstract models of repository. In this article, meta data of compound knowledge are defined by being classified into the two factors. A component stands for the property about a main agent, activity unit or resource that use and create knowledge, and a context presents the context in which knowledge object are included. An agent of the compound knowledge process performs classification, registration, and pattern information management of composite knowledge, and serves as data flow and processing between compound knowledge repository and user. The agent of the compound knowledge process consists of the following functions: warning to inform data search and extraction, data collection and output for data exchange in an distributed environment, storage and registration for data, request and transmission to call for physical material wanted after search of meta data. In this article, the construction of a compound knowledge repository for recommendation system to be developed can serve a role to enhance learning productivity through real-time visualization of timely knowledge by presenting well-put various contents to users in the field of industry to occur work and learning at the same time.

• Journal of KIISE
• /
• v.43 no.5
• /
• pp.579-589
• /
• 2016

Currently, corporations have introduced the knowledge management system that utilizes knowledge effectively for practical purpose and development of core ability. However, existing knowledge systems have failed to share the knowledge content due to lack of elements that encourage the members to participate in the system. In this study, we designed a novel knowledge management system that employs the structure of social network service (SNS). More precisely, screen layout according to function and several algorithms to improve user friendliness and produce integrated knowledge content are recommended. The proposed SNS-based knowledge management system encourages the enterprise members to participate in the system to produce and share valuable knowledge contents.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06b
• /
• pp.393-395
• /
• 2012

많은 현대인들은 체중 관리를 위해 많은 시간과 노력을 쏟고 있으며 그중에서도 식단을 관리하는데 많은 힘을 기울이고 있다. 하지만, 전문지식이 없는 일반인이 자신이 먹은 식단을 분석하고 어떤 음식을 먹을지 계획하는 것은 쉽지 않다. 따라서 본고에서는 hypernetwork를 이용한 개인 맞춤형 식단 추천 알고리즘을 제안한다. 개발된 식단 추천 알고리즘은 사용자의 식단 로그 데이터를 기반으로 사용자의 식성에 맞고 적절한 칼로리를 지닌 식단을 구성하여 추천한다. 특히, 식품 정보 DB 이외에 다른 추가 정보가 필요하지 않으며, 개인의 작은 식단 로그 데이터만으로도 동작 가능한 장점을 가지고 있다. 본 연구실에서는 개발된 알고리즘을 이용하여 개인 체중 관리 어플리케이션인 DietAdvisor를 제작하였으며, 사용자는 어플리케이션을 통해 실제 식단 추천 및 그 외의 체중관리에 필요한 서비스를 제공받을 수 있다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2011.04a
• /
• pp.1308-1309
• /
• 2011

u-헬스 컨텐츠 증가에 따라 사용자들 자신에게 적합한 정보를 찾기 위해 많은 시간이 요구된다. 본 논문에서는 사용자 프로필 정보와 분야 전문 지식을 바탕으로 사용자에게 가장 적합한 아이템을 추천해주는 시스템을 제안한다. 제안하는 방법을 통해 기존 추천 시스템에서 발생하는 overspecialization 문제와 new user/item 문제를 완화 시킬 수 있다. 실험을 통해 제안하는 추천 시스템이 u-헬스 컨텐츠 추천에 적합함을 정성적으로 보인다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2006.05a
• /
• pp.427-430
• /
• 2006

지식과 정보의 양이 급격히 증가함에 따라 일정한 기준으로 지식을 분류하기 어려울 뿐만 아니라 분류정보에 대한 지속적인 관리노력이 요구되었다. 본 논문에서는 다양한 지식 원천에서 여러 경로를 통해 지식들이 수집되는 경우에 수집된 지식들을 지식경영시스템의 지식 맵에 맞추어 분류한다. 또한 지식경영사용자 중 특정분야에 관심이 있다고 표시한 사용자나 그 분야의 전문가에게 새로 수집된 지식을 추천하게 하여 신속하게 관련지식을 제공하여 줄 수 있게 한다. 이는 사용자의 사용이력을 가지고 사용자의 전문분야를 선정하는 시스템을 구현하는 것이다. 이는 국방 분야를 비롯한 여러 분야에 활용이 가능하여 지식 분류를 위한 노력을 감소시키고 사용자의 전문분야 파악을 통해 전문분야 정보관리 가 용이해 진다.

• The Korean Publising Journal, Monthly
• /
• s.200
• /
• pp.10-19
• /
• 1996

• Journal of the Korean Geographical Society
• /
• v.44 no.3
• /
• pp.372-394
• /
• 2009

This study investigates the ontology-based public participation GIS(PPGIS). The major reason that ontology-based GIS has attracted attention in semantic communication in recent year is due to the wide availability of geographical variable and the imminent need for turning such recommendation into useful geographical knowledge. Therefore, this study has been focused on designing and implementing the pilot tested system for public participation GIS. The applicability of the pilot tested was validated through a simulation experiment for history tourism in Guri city Gyeongi-do, Focused on the methodology, the life cycle model which involves regional statues and user recognition, can be viewed as an important preprocessing step(specification, conceptualization, formalization, integration and implementation) for recommended geographical knowledge discovery by axiom. Focusing on practicality, ontology in this study would be recommended for geographical knowledge through reasoning. In addition, ontology-based public participation GIS would show integration epistemological and ontological approach, and be utilized as an index which is connected with semantic communication. The results of the pilot system was applied to the study area, which was a part of scenario. The model was carried out using axiom of logical constraint in the meaning of human-activity.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.603-605
• /
• 2022

추천시스템은 개개인의 성향에 따른 맞춤화 추천이 가능하기 때문에 음악, 영상, 뉴스 등 많은 분야에서 관심을 받고 있다. 일반적인 추천시스템 모델은 블랙박스 모델이기 때문에 추천 결과에 따른 원인 도출을 할 수 없다. 하지만 XAI 의 모델은 이러한 블랙박스 모델의 단점을 해결하고자 제안되었다. 그 중 KGAT 는 Attention Score 를 기반으로 추천 결과에 따른 원인을 알 수 있다. 이와 같은 AI, XAI 등의 딥 러닝 모델에서 각각의 활성화 함수는 상황에 따라 상이한 성능을 나타낸다. 이러한 이유로 인해 데이터에 맞는 활성화 함수를 적용해보는 다양한 시도가 필요하다. 따라서 본 논문은 XAI 추천시스템 모델인 KGAT 의 성능 개선을 위해 여러 활성화 함수를 적용해보고, 실험을 통해 수정한 모델의 성능이 개선됨을 보인다.

• Knowledge Management Research
• /
• v.21 no.3
• /
• pp.27-44
• /
• 2020

In order to solve the user's information overload problem, recommender systems infer users' preferences and suggest items that match them. The collaborative filtering (CF), the most successful recommendation algorithm, has been improving performance until recently and applied to various business domains. Visual information, such as book covers, could influence consumers' purchase decision making. However, CF-based recommender systems have rarely considered for visual information. In this study, we propose VizNCS, a CF-based deep learning model that uses visual information as additional information. VizNCS consists of two phases. In the first phase, we build convolutional neural networks (CNN) to extract visual features from image data. In the second phase, we supply the visual features to the NCF model that is known to easy to extend to other information among the deep learning-based recommendation systems. As the results of the performance comparison experiments, VizNCS showed higher performance than the vanilla NCF. We also conducted an additional experiment to see if the visual information affects differently depending on the product category. The result enables us to identify which categories were affected and which were not. We expect VizNCS to improve the recommender system performance and expand the recommender system's data source to visual information.

• Knowledge Management Research
• /
• v.21 no.4
• /
• pp.175-193
• /
• 2020

With the proliferation of information technology communication and smart device, the environment where contents are produced and distributed is changing. People can use the contents quickly and easily, and the content industry is attracting attention and creating newly added value by converging with other industries. Accordingly, there is a need for content-related companies to understand the quality of content perceived by users in order to succeed in content, and to use it strategically. Therefore, this study aims to examine the relationship between content quality factors, user satisfaction, and recommendation intention through empirical analysis based on an IS success model. The analysis was conducted using smartPLS3.0 based on a total of 301 survey responses. As a result of the study, it was found that content usefulness, accessible system quality, convenient system quality, service provider trust, and interaction had a significant effect on user's satisfaction. Perceived privacy protection had a significant effect on user satisfaction and recommendation intention. Lastly, it was found that user satisfaction had a significant effect on recommendation intention. The results of this study are expected to provide useful information and therefore content companies can understand about the quality perceived by users.