Mucheol Kim

Big-data: transformation from heterogeneous data to semantically-enriched simplified data

In big data, data originates from many distributed and different sources in the shape of audio, video, text and sound on the bases of real time; which makes it massive and complex for traditional systems to handle. For this, data representation is required in the form of semantically-enriched for better utilization but keeping it simplified is essential. Such a representation is possible by using Resource Description Framework (RDF) introduced by World Wide Web Consortium (W3C). Bringing and transforming data from different sources in different formats into the RDF form having rapid ratio of increase is still an issue. This requires improvements to cover transition of information among all applications with induction of simplicity to reduce complexities of prominently storing data. With the improvements induced in the shape of big data representation for transformation of data to form into Extensible Markup Language (XML) and then into RDF triple as linked in real time. It is highly needed to make transformation more data friendly. We have worked on this study on developing a process which translates data in a way without any type of information loss. This requires to manage data and metadata in such a way so they may not improve complexity and keep the strong linkage among them. Metadata is being kept generalized to keep it more useful than being dedicated to specific types of data source. Which includes a model explaining its functionality and corresponding algorithms focusing how it gets implemented. A case study is used to show transformation of relational database textual data into RDF, and at end results are being discussed.

Real-time imaging-based assessment model for improving teaching performance and student experience in e-learning

Abstract Multimedia is an essential and integral part of electronic learning (e-learning). In this study, teaching performance and student learning experience are measured using real-time multimedia processing tools and techniques for the e-learning paradigm. Visual attention and visual engagement analysis are performed using two developed algorithms. Video lectures are recorded and delivered to students in e-learning pedagogical setup, which are examined for the visual attention and visual engagement of the student and teacher, respectively. Proposed methodology integrates the assessment on both student and teacher ends. Multimedia processing of video lectures for teaching performance produces scoring dataset. The same methodology on student end for visual attention is used to investigate student experience. These types of datasets then reduced to time-based datasets from the image-based dataset. Correlation and association of both datasets provide the opportunity to relate both student experience and teaching performance as well as to move forward to create content that is more useful. Computational performance of the developed algorithms is compared using different video lectures with their processed frames per second, which is analyzed as per their corresponding bins. Mean, max, and median of the processed frames of all the processed videos are also compared.

Uncertain clustering algorithms based on rough and fuzzy sets for real-time image segmentation

In real pattern recognition applications, the complete and accurate information of a given set is not always easy to get. Such incomplete knowledge may lead to imperfect expressions of the set using many pattern recognition methods. Rough sets theory is designed to approximately describe an imprecise set by a pair of lower and upper approximations which are weighted by different parameters. As the distributive character varies from one set to another, it is undesirable to employ a constant weighted parameter for controlling the importance of the lower and upper approximations on describing various given sets. This paper presents an improved rough-fuzzy c-means clustering algorithm in which a parameter selection strategy is designed to adaptively adjust the weighted parameter depending on the distributive character of each cluster instead of manually choosing a constant parameter. Such an online-decision method enables the formed prototype to get close to the desirable location. Experimental results on synthetic datasets, real-life datasets, and image segmentation problems confirm the effectiveness of the proposed adaptive parameter selection strategy. With the introduction of adaptive parameter selection strategy, the improved rough sets-based clustering algorithm outperforms its counterparts in certain cases.

Context-Aware Recommendation Model based on Mobile Application Analysis Platform

Recently, various types of log data have been collected and used due to the explosive increase of mobile devices. In mobile environment with high portability and mobility, in addition, the user context information is an important factor for recommendation process. This study attempted to analyze usability log data collected from the mobile device through an application analysis platform. We suggested a context-aware recommendation model to recommend mobile applications or contents by recognizing users’ context data. The usability data of applications consist of activities which were active during the use of a mobile device. The features of these activities are related with time, location and device information. A model proposed in this study has a flexible structure which can be selectively used depending on user circumstances and performs a usability patterns of the applications based on the collaborative filtering method.

Resource management model based on cloud computing environment

In this article, we propose the dynamic resources management model in a cloud computing environment. For monitoring the certain resource, we should utilize not only a cloud management module but also a network management module. However, it is difficult to check the duration time and to observe the digested information about the resources. To investigate these problems in a cloud computing environment, we designed and deployed the cloud service infrastructure based on open-source software, namely, CloudStack. The proposed model regularly stores the usage data for computing resources based on Hadoop and HBase. In addition, our model analyzes the raw data for virtual machines and makes an effective recommendation regarding the consumption of computing resources.

Dynamic knowledge management from multiple sources in crowdsourcing environments

Due to the spread of smart devices and the development of network technology, a large number of people can now easily utilize the web for acquiring information and various services. Further, collective intelligence has emerged as a core player in the evolution of technology in web 2.0 generation. It means that people who are interested in a specific domain of knowledge can not only make use of the information, but they can also participate in the knowledge production processes. Since a large volume of knowledge is produced by multiple contributors, it is important to integrate and manage knowledge efficiently. In this paper, we propose a social tagging-based dynamic knowledge management system in crowdsourcing environments. The approach here is to categorize and package knowledge from multiple sources, in such a way that it easily links to target knowledge.

Multimedia based student-teacher smart interaction framework using multi-agents in eLearning

Multimedia content comprises the graphics, audio & video clips, animation and text to present learning materials in a style, which improves learner expectation in eLearning paradigm. Electronic learning gained the popularity due to its immense coverage of students and subjects all over the world. The aim of this study is enhancements using agent-based framework through multimedia data in eLearning paradigm. Analysis of multimedia contents and eLearning data are helpful for the course designers, teachers, and administrators of eLearning environments to hunt for undetected patterns and underlying data in learning processes. This research improves the learning curves for the students. It also needs to improve the overall processes in eLearning paradigm. Information and Communication Technologies supported education, and virtual classrooms environments are mandatory. In eLearning data is evolving day by day that includes the semi-structured data, unstructured data, and structured data which is also collectively marked as multimedia big data. Multimedia data has the potential to mining for the analytics and learning. The learning outcomes for the students are very important to find the facts that what impacts the input data on the student. There are 1108 students posted questions in online Learning Management System (LMS) and instructors reply these queries. Sensor data is also gathered by the mobile GPS to find the student location. The system has analyzed the relevance of the replied answers. The student satisfaction is achieved by providing the multimedia-based student-teacher interaction. This can lead to synchronous communication and multimedia content conversation in eLearning paradigm. Machine learning techniques are applied to that data to discover the patterns and behavioral trends. It can also be used in the eLearning environments for the teacher to assist and enhance the pedagogical skills and for student’s learning curve enhancements.

Crowdsourcing based Scientific Issue Tracking with Topic Analysis

Abstract With the advancement of web technologies, many people are participating in the information production and distribution process in the Web environment. In addition, many researchers have been interested in research on refining useful information using topic based recommendation system because the amount and complexity of web information is rapidly increasing. The proposed approach performs typical scientific data collection and then analyzes seed problem keywords using multi-level documents based on crowd sourcing. We then used the LDA algorithm to create a cluster of scientific themes to generate issue keywords that are responsive to the scientific trend issues. As a result, our approach suggests a methodology for recommending clusters of related issues when scientific issues are raised in each context.

Impact of Dynamic Path Loss Models in an Urban Obstacle Aware Ad Hoc Network Environment

This study highlights the importance of the physical layer and its impact on network performance in Mobile Ad Hoc Networks (MANETs). This was demonstrated by simulating various MANET scenarios using Network Simulator-2 (NS-2) with enhanced capability by adding propagation loss models (e.g., modified Two-Ray Ground model, ITU Line of Sight and Nonline of Sight (ITU-LoS and NLoS) model into street canyons and combined path loss and shadowing model (C-Shadowing)). The simulation results were then compared with the original Two-Ray Ground (TRG) model already available into NS-2. The scenario primarily simulated was that of a mobile environment using Random Way Point (RWP) mobility model with a variable number of obstacles in the simulation field (such as buildings, etc., causing variable attenuation) in order to analyze the extent of communication losses in various propagation loss models. Performance of the Ad Hoc On-demand Distance Vector (AODV) routing protocol was also analyzed in an ad hoc environment with 20 nodes.

Multimedia Management Services Based on User Participation with Collaborative Tagging

As Internet technology has rapidly developed, the amount of multimedia content on the Web has expanded exponentially. Collaborative tagging, namely folksonomy, is emerging to promote user participation in generating and distributing active content. This could be significant evidence for categorizing dynamic multimedia content. For that reason, we proposed an efficient multimedia management system based on collaborative tagging. Our system suggests the candidates, with collaborative filtering for describing and categorizing the multimedia content.