Seong Baeg Kim

Measuring similarity of android applications via reversing and K-gram birthmarking

By measuring similarity of programs, we can determine whether someone illegally copies a program from another program or not. If the similarity is significantly high, it means that a program is a copy of the other. This paper proposes three techniques to measure similarity of the Dalvik executable codes (DEXs) in the Android application Packages (APKs). Firstly, we decompile the DEXs of candidate applications into Java sources and compute the similarity between the decompiled sources. Secondly, candidate DEXs are disassembled and the similarities between disassembled codes are measured. Finally, we extract k-gram based software birthmark form the dissembled codes and calculate the similarity of sample DEXs by comparing the extracted birthmarks. We perform several experiments to identify effects of the three techniques. With the analysis of the experimental results, the advantages and disadvantages of each technique are discussed.

Monitoring-based temporal prediction of power entities in smart grid cities

This paper develops prediction models for groundwater and wind speed in smart grid cities, where the gradual penetration of electric vehicles keeps increasing load in its power system. As a major consumption source and a renewable energy source, water management systems and wind power generators make operation plans for their facilities based on temporal predictions of target objects. The massive data of day-by-day groundwater level readings and day-by-day wind speed records, which have been accumulated for about 10 years in Jeju City, are fed to 3-layer neural networks for modeling the change pattern. The trained neural network yields an accurate short-term prediction for groundwater level, with its maximum error less than 4.76 %. In addition, for wind speed predictions, 40 % of forecast errors are smaller than 5 %, in spite of its error spikes reaching 37 %. Those models are provided to the electric vehicle charging system to shift the peak load to less loaded intervals.

Improvement of Jeju National Museum service using mobile RFID

Recently, various types of the museum services using RFID technology have been developed as RFID technology evolves. However, most of them are yet limited to transfer the simple information of exhibits or manage the exhibits of museums warehouse. With reflection on the island features of Jeju National Museum, we have developed a system, which can improve the service quality of the museum by providing the disruptive services using mobile RFID technology. The system we developed accumulates the information such as the number of viewing an exhibit, viewers age, and viewers gender, etc., whenever a viewer with an RFID reader connected to a PDA reads a tag on an exhibit. This information is directly stored in a database, which can provide an efficient search service, to bring better service quality of the museum. We have implemented the application program for the viewers pattern analysis based on the data retrieved from the database. Using this program, we can get the movement paths of the viewers. Also, we propose an approach to bring real-time customized interaction between a viewer and a curator to answer an intriguing question of a viewer on the spot where an exhibit is located. We expect that our approach proposed will be a disruptive model widely used for most museums because it can improve their services in spite of reducing the cost of their operation and management.

The Relative Importance of Weather Factors and the Predictions About the Groundwater Level in Jeju

The research related to the groundwater level in Jeju has focused on the amount of rainfall and its hydrogeological characteristics so far. However, since the sensor technology has been used in many areas to measure the various types of natural phenomena recently, it allows us to perform advanced analysis on the groundwater in Jeju. In this paper, we consider wind speed, evaporation, temperature, humidity, and rainfall as the factors that can have influences on the groundwater level. We describe how the factors can affect the groundwater and how artificial neural networks can predict the groundwater level. We perform multiple regression and hierarchical linear model analysis. And then, we calculate the relative importance of the weather factors. We use data accumulated from 2003 to 2009 and perform data-oriented analysis rather than theoretical analysis. We divide Jeju region into four basins such as north (Jeju), south (Seogwipo), west (Gosan), and east (Sungsan).

Mobile Digital Storytelling Development for Energy Drink’s Promotion and Education

Digital storytelling is the process in which stories are expressed in digital media, which is invaluable due to the limitless potential in the field of advertisement and education. An intriguing digital storytelling is to have engaging and high sensual stimulating animations that contain the narration required for the desired result. For this study, we deal with energy drinks which have recently had a high attention and rapidly growing markets. In particular, we focus on Jeju specific energy drinks along with a narration for hand-held devices such as mobile phones by implementing interactive animations which include a touch interface and a narration involved.

Design of a slot assignment scheme for link error distribution on wireless grid networks

This paper designs and measures the performance of an efficient routing scheme and corresponding slot assignment schedule capable of efficiently handling the difference in the error rate of each link on the wireless mesh network Targeting at the grid topology common in most modern cities, a split-merge operation masks the channel error by making an intermediary node receive a message simultaneously in two directions and then integrating a secondary route, required to have the same length with the primary route, to the communication schedule Adding virtual links from level 1 to level 3 to the 3 × 3 grid and estimating their error rates based on those of the surrounding links, the proposed scheme applies the well-known shortest path algorithm to decide the best route and also to finalize the complete a slot assignment for the WirelessHART protocol The performance measurement result obtained by simulation using a discrete event scheduler demonstrates that the proposed scheme can find the promising path, improving the delivery ratio by up to 5.21 %.

A diagnostic model using a clustering scheme

It has been recognized that it is a challenging problem to deal with the situation where learners have diverse computing backgrounds and the learning content to be covered is also in the broad coverage. In the case, it’s required to devise a sophisticated diagnostic model for applying a proper teaching-learning method. We have drawn a scheme which can be applied to that case efficiently by using clustering algorithms based on web technology. In our approach, we focus on finding out methods for classifying both learners and learning content on the web. To make classification and manipulation of learning content ease, we reorganize learning content in order to have discrete form by introducing the concept of the knowledge unit which is extracted from each topic. Also, to make classification and diagnostic ease, we develop questions to measure them and analyze each question using item response theory (IRT) on the web. From the experiment of students sampled using our method, we show that learners with various backgrounds and the learning content with distribution on the broad range can be categorized effectively into the groups with homogeneous property. Also, we describe how to apply our proposed scheme to the introductory courses at postsecondary level.

A profiling system and mobile middleware framework for u-learning

With recently increased emphasis on learner-centered education, a u-learning system, which can provide learners with a learning at any time in any place, has been becoming the focus of attention. Also, the advance of technology enables learners to have easier access to internet through PDA, thus to acquire information needed to learn on the spot at the moment without time or space constraints. Based on the current trend of the technology, we focus on how learners obtain and access information suitable for them. We shall propose a mobile middleware framework and profiling system, which take into account circumstances of a learner with a mobile device, specifications on a mobile device, and the learners level or preferences for individualized u-learning. The middleware system is devised to make use of the information collected from RFID reader for outlining the learning environment of the learner.

App Development for Jeju Spring Water Information Visualization

A new trend, which tries to apply an information technology to the field of an existing or emerging industry, is becoming prevalent with the rapid advent of information technology. This study is closely related with an information technology (IT) fusion research on Jeju spring water, aiming at the management of Jeju water resources and smart appliances app service of Jeju spring water information, combined with IT. Using smart appliances, we provide such information as Jeju spring water location information, the surrounded information, and the information on how to use it. It will help to make people recognize and promote the high-quality of Jeju spring water. We focus on how to apply key features of smart appliances to spring water information visualization and management system. Furthermore, it can be applied as an important reference in the development and management of Jeju spring water.

A Novel Approach to Provide Mobile RFID-Based Services Combined with Data Mining Technique

In advance of RFID technology, various types of services based on mobile RFID technology have been developed. However, there still exists limitation on RFID-based services such as transferring the simple information of exhibits at exhibition halls or managing the exhibits stored in warehouses. In order to improve the service qualities at exhibition centers, showrooms, or museums, we have developed a system, which can provide the disruptive services using mobile RFID technology. Our system accumulates the information such as how many people visit a specific exhibit, what the demographic information the visitors have, and what kinds of items the visitors see, and so on whenever visitors read a tag on an exhibit with an RFID reader connected to a mobile device. The accumulated information is stored in a database to provide an efficient search service and to bring a better service to the visitors. We have developed an application program for analyzing the visitors’ data retrieval pattern from the database. Using this program, our system can get the information about the movement paths of the visitors. Also, by adopting data mining technology, we also propose a novel approach to enhance RFID-based services so that visitors can have a customized real-time question-answering service from assistants in a certain exhibition hall or museum whenever they are intrigued to know about exhibits.