Aekyung Moon

Designing CAMUS based Context-Awareness for Pervasive Home Environments

With the advent of pervasive computing environments, the issue related to context-aware middleware is an emerging area of research. We propose proactive services on the basis of CAMUS (context-aware middleware for URC Systems). First, this paper presents the system architecture of CAMUS. CAMUS is a middleware for providing context-aware applications with development and execution methodology. Accordingly, the applications developed by CAMUS respond in a timely fashion to contexts. Secondly, this paper illustrates the contents recommendation and control service agents with the property, operation, and task for the context-aware active services. Next, proposed active services is applied to the pervasive home environment for evaluation of our system. To do this, we implement the TV contents recommendation service agent, control service agent and context-aware task using CAMUS

An Efficient Context Model for Fast Responsiveness of Context-Aware Services in Mobile Networks

In order to provide context-aware services, context modeling is one of the most important tasks. Ontology has been widely accepted for the context modeling but it brings some overheads in the process of reasoning large-scale and dynamic information. In mobile networks, especially, ontology should be verified to be applicable for Context-aware Services since real-time responsiveness for large scale moving users is highly required to be guaranteed. This paper proposes an efficient context model to overcome the limitation of ontology. For the purpose, we design a conceptual context model and compare the processing time of the context-aware mobile service system implying the three different context modeling approaches - ontology, database, and proposed hybrid modeling. As the result of performance evaluations, the proposed hybrid model shows faster processing time among them and lesser affected by the major metrics of mobile services - changeability and number of users - than the two existing models. Based on the experimental results, we expect that the proposed hybrid model is very useful and efficient for fast responsiveness to support Context-aware Services for large scale users in mobile networks.

Context-aware user model for personalized services

With the advent of ubiquitous computing environments, a significant amount of work has been devoted to context-aware. We proposed the context-aware user model, which is aware of its userpsilas situation and recommends personalized services based on this information. This paper focuses on how our algorithms can be applied to a variety of environment easily and evolve user model without previously known information since the concerns of privacy. The proposed user model consists of context and action pairs. Using context-aware user model, we learn user behavior pattern in the situation, which is captured by union of contexts according to current environments. In order that our user model applies to new environment, only context profiles are required to be defined without any modification to model. Moreover we evolved user model through interactions between users and mobile devices using learning mechanism. Finally, we evaluate the precision of proposed approach using simulation with data sets of UCI depository.

Ubiquitous Plant Farming based on networks

Ubiquitous Plant Farming based on network is a service which is capable of coping with environmental and technical problems with the actualized IT convergence case for agriculture. This paper presents a brief survey on the standardization status including the reference model, service requirements and network capabilities prepared for this purpose.

Parallel hash join algorithms for dynamic load balancing in a shared disks cluster

Most of previous parallel join algorithms assume a shared nothing (SN) cluster, where each database partition is owned by a single processing node. While SN cluster can interconnect a large number of nodes and support a geographically distributed environment, it may suffer from poor facility for load balancing and system availability compared to a shared disks sharing (SD) cluster. In this paper, we first propose a dynamic load balancing strategy by exploiting the characteristics of SD cluster. Then we parallelize conventional hash join algorithms using the dynamic load balancing strategy. We also explore the performance of parallel join algorithms using a simulation model of SD cluster. The experiment results show that the proposed parallel join algorithms can achieve higher potential for dynamic load balancing with the inherent flexibility of SD cluster.

Global Concurrency Control Using Message Ordering of Group Communication in Multidatabase Systems

A multidatabase system (MDBS) is designed to provide universal access to distributed data across multiple autonomous and possibly heterogeneous local database systems. In this paper, we propose a new global concurrency control algorithm, named GCC-M, in the MDBS. GCC-M is the first algorithm that integrates the idea of message ordering of group communication to the global concurrency control. The message ordering allows to determine easily the relative serialization order of global transactions, and thus it can make the global concurrency control algorithm be simple and can also avoid distributed deadlocks. Using a distributed database simulation model, we show that GCC-M outperforms the previous global concurrency control algorithms in a wide variety of database workloads.

Components Based Integrated Management Platform for Flexible Service Deployment in Plant Factory.

The plant factory is a facility that aid growers to make the steady production of high-quality vegetables and ornamental plants all year round by artificially controlling the cultivation environment such as indoor temperature and humidity. There are various types of plant factory such as fully artificial light-type, sunlight-type and plastic house, etc. In this paper, we propose Intelligent Greenhouse Integrated Management Platform (IGIMP) that can support a variety of plant factory type with the flexible service deployment. To optimize the environment for crop growth automatically, it gathers environmental context information from various indoor sensors as well as outdoor sensors and provides appropriate control functions using the convergence of ubiquitous computing and agriculture technologies. Finally, we introduce implementation through small plant factory prototype a kind of fully artificial light-type greenhouse.

Higher order knowledge management platform for the personalized services in next generation networks

We propose higher order knowledge management platform (HKMP) that can support to provide the personalized services for the end users by means of evolving lower order network knowledge. HKMP helps 3rd party service providers to create personalized services considering users context through the open interfaces for providing higher order knowledge. In this paper, we classify lower order network knowledge and build ontology to represent them including user profiles and users preference. Furthermore, we propose the efficient algorithms of leaner and recommender to evolve higher order knowledge in HKMP. Finally, we simulated to evaluate the precision of proposed algorithms with data sets of UCI depository. As the result of evaluation, we expect that the HKMP would be an essential component for providing personalized services in next generation networks.

Performance of dynamic load balanced join algorithms in shared disk parallel database systems

Parallel database systems are the key to high performance database processing. In this paper, we propose parallel join algorithms in shared disk parallel database systems, where all coupled nodes are connected via a high-speed network and share a common database at the disk level. The proposed algorithms are novel in the sense that they can provide a higher potential for dynamic load balancing with the inherent flexibility of the shared disk architecture. Using a parallel database simulation model, we evaluate the performance of the proposed algorithms under a wide variety of system configurations and database workloads.

Disease and pest prediction IoT system in orchard: A preliminary study

Each species of diseases and pests is considered to be harmful to plants and has a great adverse effect on agriculture. In addition to decreasing crop productivity and quality, pesticides or fungicides which cause environmental pollution are needed to kill diseases and pests. Therefore, the IoT system is designed to reduce the frequent use of insecticides and fungicides and to predict when the pests appear in order to lower the appearance of pests. Therefore, weather stations near the orchard have been installed and used to analyze the correlation between pests and weather data that had a large effect on pests. The prediction models have been created utilizing correlation information. In other words, we proposed a system that provides disease and pest prediction information so that farmers can quickly control them.