• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.40 no.4
• /
• pp.154-163
• /
• 2017

Collaborative filtering, one of the most widely used techniques to build recommender systems, is based on the idea that users with similar preferences can help one another find useful items. Credit card user behavior analytics show that most customers hold three or less credit cards without duplicates. This behavior is one of the most influential factors to data sparsity. The 'cold-start' problem caused by data sparsity prevents recommender system from providing recommendation properly in the personalized credit card recommendation scenario. We propose a personalized credit card recommender system to address the cold-start problem, using multiple user profiles. The proposed system consists of a training process and an application process using five user profiles. In the training process, the five user profiles are transformed to five user networks based on the cosine similarity, and an integrated user network is derived by weighted sum of each user network. The application process selects k-nearest neighbors (users) from the integrated user network derived in the training process, and recommends three of the most frequently used credit card by the k-nearest neighbors. In order to demonstrate the performance of the proposed system, we conducted experiments with real credit card user data and calculated the F1 Values. The F1 value of the proposed system was compared with that of the existing recommendation techniques. The results show that the proposed system provides better recommendation than the existing techniques. This paper not only contributes to solving the cold start problem that may occur in the personalized credit card recommendation scenario, but also is expected for financial companies to improve customer satisfactions and increase corporate profits by providing recommendation properly.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.34 no.6
• /
• pp.1873-1879
• /
• 2014

There are various methodologies to forecast the travel time using real-time data but the K-nearest neighborhood (KNN) method in general is regarded as the most one in forecasting when there are enough historical data. The objective of this study is to evaluate applicability of KNN method. In this study, real-time and historical data of toll collection system (TCS) traffic flow and the dedicated short range communication (DSRC) link travel time, and the historical path travel time data are used as input data for KNN approach. The proposed method investigates the path travel time which is the nearest to TCS traffic flow and DSRC link travel time from real-time and historical data, then it calculates the predicted path travel time using weight average method. The results show that accuracy increased when weighted value of DSRC link travel time increases. Moreover the trend of forecasted and real travel times are similar. In addition, the error in forecasted travel time could be further reduced when more historical data could be available in the future database.

• Journal of Internet Computing and Services
• /
• v.10 no.4
• /
• pp.1-11
• /
• 2009

Recently there has arisen concern in both the database community and the graphics society about data retrieval from large motion databases because the high dimensionality of motion data implies high costs. In this circumstance, finding an effective distance measure and an efficient query processing method for such data is a challenging problem. This paper presents an elaborate motion query processing system, SMoFinder (Similar Motion Finder), which incorporates a novel kinematic distance measure and an efficient indexing strategy via adaptive frame segmentation. To this end, we regard human motions as multi-linkage kinematics and propose the weighted Minkowski distance metric. For efficient indexing, we devise a new adaptive segmentation method that chooses representative frames among similar frames and stores chosen frames instead of all frames. For efficient search, we propose a new search method that processes k-nearest neighbors queries over only representative frames. Our experimental results show that the size of motion databases is reduced greatly (${\times}1/25$) but the search capability of SMoFinder is equal to or superior to that of other systems.

• The Journal of Korean Association of Computer Education
• /
• v.5 no.3
• /
• pp.37-44
• /
• 2002

Automatic text classification is a task of assigning predefined categories to free text documents. Its importance is increased to organize and manage a huge amount of text data. There have been some researches on automatic text classification based on machine learning techniques. While most of them was focused on proposal of a new machine learning methods and cross evaluation between other systems, a through evaluation or optimization of a method has been rarely been done. In this paper, we propose an improving method of kNN-based Korean text classification system using heuristic informations about decision function, the number of nearest neighbor, and feature selection method. Experimental results showed that the system with similarity-weighted decision function, global method in considering neighbors, and DF/ICF feature selection was more accurate than simple kNN-based classifier. Also, we found out that the performance of the local method with well chosen k value was as high as that of the global method with much computational costs.

• Journal of KIISE
• /
• v.42 no.12
• /
• pp.1611-1622
• /
• 2015

Although Android systems provide a permission-based access control mechanism and demand a user to decide whether to install an app based on its permission list, many users tend to ignore this phase. Thus, an improved method is necessary for users to intuitively make informed decisions when installing a new app. In this paper, with regard to the permission-based access control system, we present a novel approach based on a machine-learning technique in order to support a user decision-making on the fly. We apply the K-NN (K-Nearest Neighbors) classification algorithm with necessary weighted modifications for malicious app classification, and use 152 Android permissions as features. Our experiment shows a superior classification result (93.5% accuracy) compared to other previous work. We expect that our method can help users make informed decisions at the installation step.

• Bulletin of the Korean Chemical Society
• /
• v.5 no.3
• /
• pp.117-121
• /
• 1984

The structure of $CS_{7.3}Ag_{4.7}Si_{12}Al_{12}O_{48}$, vacuum dehydrated zeolite A with all Na+ ions replaced by $Cs^+$ and $Ag^+$ as indicated, has been determined by single-crystal x-ray diffraction techniques in the cubic space group, Pm3m (a = 12.282 (1) ${\AA}$). The structure was refined to the final error indices $R_1$$R_2$ (weighted) = 0.099 using 347 independent reflections for whind intlch $I_0\;>\;3{\sigma}(I_0)$. Although deydration occurred at $360^{\circ}C$, no silver atoms or clusters have been observed. The 8-ring sites are occupied only by $Cs^+$ ion, and the 4-ring sites only by a single $Ag^+$ ion. The 6-ring sites contain $Ag^+$ and $Cs^+$ ions with $Ag^+$ nearly in 6-ring planes and $Cs^+$ well off them, one on the sodalite unit side. With regard to the 6-rings, the structure can be represented as a superposition of two types of unit cells: about 70 % have $4Ag^+$ and $4Cs^+$ ions, and the remaining 30 % have $3Ag^+$ and $5Cs^+$. In all unit cells, $3Cs^+$ ions lie at the centers of the 8-rings at sites of D4h symmetry; these ions are approximately 0.3 ${\AA}$ further from their nearest framework-oxygen neighbors than the sum of the appropriate ionic radii would indicate. To minimize electrostatic repulsions, the $Cs^+$ ions at Cs(1) are not likely to occupy adjacent 6-rings in the large cavity; they are likely to be tetrahedrally arranged when there are 4.