Won-Ik Park

A Context-Based Collaboration System in Ubiquitous Environments

In ubiquitous environments, invisible devices and software are connected to one another to provide convenient services to users. In order to provide such services, we must have mobile devices that connect users and services. However, the types of services available have thus far been limited due to the limited resources of mobile devices. This paper will propose a solution to the resource problem of mobile devices by presenting a context-based collaboration system that allows mobile devices to share various nearby resources. We have tested our context-based collaboration system by using the test scenario. This paper focuses on sharing different devices by recognizing context data in a given situation.

A split synchronizing mobile transaction model

We propose a new notion of hoardset fragmentablilty and SSMT(Split Synchronizing Mobile Transaction) model for ubiquitous mobile clients, which not only assures transaction atomicity in synchronization but also reduces synchronization cost and provides context-awareness in ubiquitous computing environment with uneven conditioning. It splits the synchronization process of a mobile transaction into 2 parts by exploiting the hoardset fragmentability of data objects which the mobile transaction access: preferred and deferred synchronization part. The deferred synchronization part which has no possibility of conflict uses the data-centric synchronization strategy and its synchronization is delayed. The preferred synchronization part with possibility of conflicts is synchronized early using the transaction-centric synchronization strategy. We also propose a new data hoarding and synchronization mechanism which facilitates the split synchronization of SSMT. The performance evaluation model for the SSMT is developed and its results are compared to a conventional transaction-centric synchronization model.

Two-phase mobile transaction validation in wireless broadcast environments

In this paper, we propose a two-phase mobile transaction validation protocol making the best use of the strong point of (l, m) indexing broadcast data organization in wireless broadcast environments. It is based on OCC(Optimistic Concurrency Control) protocol which minimizes the excessive bi-directional communication by making transactions execute unhindered to their validation point. In the proposed protocol, the validation of a mobile transaction is performed in two-phase: partial validation phase and final validation phase. The partial validation is performed at the mobile client based on the control information contained in the broadcast disk and the final validation is performed on the committed and active transactions at the server. The proposed protocol tries to minimize the restart rate of mobile transactions as well as average transaction response time through adjusting serialization order of mobile transactions by exploiting the potential read/write set of mobile transactions.

An efficient context-aware personalization technique in ubiquitous environments

It is not easy to provide a personalized service by using an existing recommendation system in ubiquitous computing environments providing the personalized service according to the situation. Because majority of existing recommendation systems focus on similarity of users and items. Although they are modified to consider the situation, they still have problems requiring more increased complexity in order to provide personalized services considering dynamically changing users situation and preference. In this paper, we propose an efficient context-aware personalization technique considering the situation and tastes of users in ubiquitous computing environments. Our method is a hybrid personalization technique successively applying context-aware personalization and user profile based personalization. And we build a context model need for the context-aware personalization. In the experimental section, we show that our method has lower complexity than traditional method through the evaluation of response time and satisfaction score.

Personalized recommender system for resource sharing based on context-aware in ubiquitous environments

In ubiquitous environments, user wants to get a variety of services using only his individual device. For this requirement, resource sharing and its efficient processing is one of the new research topics. However, the situations of each user are different and user preferences also are various even if user environments are the same. Therefore, although users want to get the same service in the same space, the most suitable resources for the service are different for all users. In this paper, we propose the personalized resource recommender system which recommends resources inferred by applying user preference and situation. The recommender system uses ontology-based reasoning and rule-based reasoning for being aware of situation and for recommending personalized resources respectively.

A Hybrid Cache Cohrency Scheme for Ubiquitous Mobile Clients

We propose a new cache coherency scheme named HCCS(hybrid cache coherency scheme) to facilitate ubiquitous data access of mobile clients. To cope with the limitations of ubiquitous computing environments, it integrates the cache invalidation scheme using broadcast channel into the publish/subscribe-based data synchronization scheme using two-way communication channel. We present novel algorithms and system architecture for the HCCS. By integrating two schemes, the HCCS provides substantial benefits: allows updates on the mobile client which are restricted in the cache invalidation scheme using broadcast channel and disconnected operations, improves the scalability. The HCCS scheme also gives proper support to the short disconnection of mobile clients, dynamically adapts to changes of disconnection period of the mobile clients.

A Personalized Device Recommender System in Ubiquitous Environments

Ubiquitous computing enhances computer use by making many computers available throughout the physical environment, while making them effectively invisible to the user. Invisible devices and software are connected to one another to provide convenient services to users. In this environment, users want to get a variety of services by using only an individual mobile device which has limitations such as tiny display screens, limited input, and less powerful processors. In other words, users hope to obtain a personalized service composed of user-suitable devices among sharable devices in a ubiquitous smart space which is called USS in this paper. However, the situations of each user are different and user preferences also are various. Therefore, although users request the same service in the same USS, the most suitable devices for composing the service are different for all users. For satisfying these user requirements, this paper proposes a device recommender system which infers and recommends user-suitable devices for composing a user required service. The objective of this paper is the development of the systems for recommending devices through context-aware inference in peer-to-peer environments. For this goal, this paper considers the environmental context and user preference Also we implement a prototype system and test performance on real ubiquitous mobile object (UMO).

Personalized digital e-library service using users’ profile information

We propose a personalized digital E-library system using a collaborative filtering technique, which provides a personalized search list according to users’ preference. The proposed system analyzes the registered users’ actions such as “clicking” and “borrowing” items. According to the different actions, we provide a weight for calculating the users’ preference of each item. However, the list is uniformly provided to the individual users when they search with same keywords. In order to avoid the problem, we customize the order of items in the list according to whether there is any mismatching of profiles among registered users and target users or not.

An agent-based personalized multimedia Jukebox for mobile devices using consumption sentiment 2

This paper classifies the users purchase pattern into four types of behavior, and proposes an agent-based personalized multimedia Jukebox that provides a personalized service on the client side. The proposed Jukebox assists the efficient management of numerous multimedia contents stored in a personal desktop computer, and provides personalized multimedia contents to the mobile device based on a push service type.

Dynamically Personalized Music Recommendation System for PDA

We introduce a novel personalized music service system through PDA. By providing only users’ preferred web pages or smaller readable service pages, the problem of the limitation of resource of PDA can be solved. In this paper, the preferred service pages are obtained from the statistical preference transactions among web pages for each web site. In computing the preference, we consider the ratio of the different of time which was visited and users’ real listening time of music recommended. We present a novel algorithm to compute the time weights for different items in a manner that will assign a decreasing weight to old data. Also, our system dynamically provides the personalized music service according to the different three cases such as the beginning stage, the positive feedback, and the negative feedback. In the experimental section, we demonstrate our personalized music service system and show how much the resource of mobile devices can be saved.