• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.12
• /
• pp.3836-3854
• /
• 2022

The student group division is benefit for universities to do the student management based on the group profile. With the widespread use of student smart cards on campus, especially where students living in campus residence halls, students' daily activities on campus are recorded with information such as smart card swiping time and location. Therefore, it is feasible to depict the students with the daily activity data and accordingly group students based on objective measuring from their campus behavior with some regular student attributions collected in the management system. However, it is challenge in feature representation due to diverse forms of the student data. To effectively and comprehensively represent students' behaviors for further student group division, we proposed to adopt activity data from student smart cards and student attributes as input data with taking account of activity and attribution relationship types from different perspective. Specially, we propose a novel student group division method based on a multi-view student attribute heterogeneous information network (MSA-HIN). The network nodes in our proposed MSA-HIN represent students with their multi-dimensional attribute information. Meanwhile, the edges are constructed to characterize student different relationships, such as co-major, co-occurrence, and co-borrowing books. Based on the MSA-HIN, embedded representations of students are learned and a deep graph cluster algorithm is applied to divide students into groups. Comparative experiments have been done on a real-life campus dataset collected from a university. The experimental results demonstrate that our method can effectively reveal the variability of student attributes and relationships and accordingly achieves the best clustering results for group division.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.10 no.4
• /
• pp.11-23
• /
• 2011

The objective of this paper is to develop a daily pattern clustering algorithm using historical traffic data that can reliably detect under various traffic flow conditions in urban streets. The developed algorithm in this paper is categorized into two major parts, that is to say a macroscopic and a microscopic points of view. First of all, a macroscopic analysis process deduces a daily peak/non-peak hour and emphasis analysis time zones based on the speed time-series. A microscopic analysis process clusters a daily pattern compared with a similarity between individuals or between individual and group. The name of the developed algorithm in microscopic analysis process is called "Two-step speed clustering (TSC) algorithm". TSC algorithm improves the accuracy of a daily pattern clustering based on the time-series speed variation data. The experiments of the algorithm have been conducted with point detector data, installed at a Ansan city, and verified through comparison with a clustering techniques using SPSS. Our efforts in this study are expected to contribute to developing pattern-based information processing, operations management of daily recurrent congestion, improvement of daily signal optimization based on TOD plans.

• The Journal of the Korea Contents Association
• /
• v.17 no.7
• /
• pp.182-192
• /
• 2017

Recently, there is an increasing interest in analysis of big data that is coming from manufacturing industry. In this paper, we use PCB (Printed Circuit Board) manufacturing data to provide manufacturers with information on areas with high PCB defect rates, and to visualize them to facilitate production and quality control. We use the K-means and DBSCAN clustering algorithms to derive the high fraction of PCB defects, and compare which of the two algorithms provides more accurate results. Finally, we develop a system of MVC structure to visualize the information about bad clusters obtained through clustering, and visualize the defected areas on actual PCB images.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.5
• /
• pp.616-620
• /
• 2010

Generally, most people use the photo management method which clusters lots of photos into each folders according to photo shooting time and date. However, since the number of photos to manage is getting more increasing, it takes much time and burdensome work. This paper describes PHOTOLAND, a system that visualizes hundreds of photos on a 2D grid space to help users manage their photos. It closely places similar photos in the grid based on temporal and spatial information. Most photograph management systems use a scrollable view based on a sequential grid layout that arranges the thumbnails of photos in some default order on the screen. Our system decreases drag and drop mouse interaction when they classify their photos into small groups comparing to the sequential grid layout. We conducted experiments to evaluate temporal coherence and space efficiency.

• The Bulletin of The Korean Astronomical Society
• /
• v.39 no.2
• /
• pp.89-89
• /
• 2014

While the current consensus view holds that galaxy mergers are commonplace, it is sometimes speculated that Globular Clusters (GCs) may also have undergone merging events, possibly resulting in massive objects with a strong metallicity spread such as Omega Centauri. Galaxies are mostly far, unresolved systems whose mergers are most likely wet, resulting in observational as well as modeling difficulties, but GCs are resolved into stars that can be used as discrete dynamical tracers, and their mergers might have been dry, therefore easily simulated with an N-body code. It is however difficult to determine the observational parameters best suited to reveal a history of merging based on the positions and kinematics of GC stars, if evidence of merging is at all observable. To overcome this difficulty, we investigate the applicability of supervised and unsupervised machine learning to the automatic reconstruction of the dynamical history of a stellar system. In particular we test whether statistical clustering methods can classify simulated systems into monolithic versus merger products. We run direct N-body simulations of two identical King-model clusters undergoing a head-on collision resulting in a merged system, and other simulations of isolated King models with the same total number of particles as the merged system. After several relaxation times elapse, we extract a sample of snapshots of the sky-projected positions of particles from each simulation at different dynamical times, and we run a variety of clustering and classification algorithms to classify the snapshots into two subsets in a relevant feature space.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.12
• /
• pp.2520-2528
• /
• 2009

In this paper, we proposed tracking detection methodology using information granulation-based fuzzy radial basis function neural networks (IG-FRBFNN). According to IEC 60112, tracking device is manufactured and utilized for experiment. We consider 12 features that can be used to decide whether tracking phenomenon happened or not. These features are considered by signal processing methods such as filtering, Fast Fourier Transform(FFT) and Wavelet. Such some effective features are used as the inputs of the IG-FRBFNN, the tracking phenomenon is confirmed by using the IG-FRBFNN. The learning of the premise and the consequent part of rules in the IG-FRBFNN is carried out by Fuzzy C-Means (FCM) clustering algorithm and weighted least squares method (WLSE), respectively. Also, Hierarchical Fair Competition-based Parallel Genetic Algorithm (HFC-PGA) is exploited to optimize the IG-FRBFNN. Effective features to be selected and the number of fuzzy rules, the order of polynomial of fuzzy rules, the fuzzification coefficient used in FCM are optimized by the HFC-PGA. Tracking inference engine is implemented by using the LabVIEW and loaded into embedded system. We show the superb performance and feasibility of the tracking detection system through some experiments.

• Journal of Environmental Impact Assessment
• /
• v.18 no.6
• /
• pp.401-409
• /
• 2009

Multivariate statistical approaches to classify sampling sites with measuring their similarity by water quality data and understand the characteristics of classified clusters have been discussed for the optimal water quality monitering network. For empirical study, data of two years (2005, 2006) at the 9 sampling sites with the combination of 2 depth levels and 7 important variables related to water quality is collected in Yongdam reservoir. The similarity among sampling sites is measured with Euclidean distances of water quality related variables and they are classified by hierarchical clustering method. The clustered sites are discussed with principal component variables in the view of the geographical characteristics of them and reducing the number of measuring sites. Nine sampling sites are clustered as follows; One cluster of 5, 6, and 7 sampling sites shows the characteristic of low water depth and main stream of water. The sites of 2 and 4 are clustered into the same group by characteristics of hydraulics which come from that of main stream. But their changing pattern of water quality looks like different since the site of 2 is near to dam. The sampling sites of 3, 8, and 9 are individually positioned due to the different tributary.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2010.05a
• /
• pp.957-961
• /
• 2010

Multivariate statistical approaches to classify sampling sites with measuring their similarity by water quality data. For empirical study, data of two years at the 9 sampling sites with the combination of 2 depth levels and 7 important variables related to water quality is collected in reservoir. The similarity among sampling sites is measured with Euclidean distances of water quality related variables and they are classified by hierarchical clustering method. The clustered sites are discussed with principal component variables in the view of the geographical characteristics of them and reducing the number of measuring sites. Nine sampling sites are clustered as follows; One cluster of 5, 6, and 7 sampling sites shows the characteristic of low water depth and main stream of water. The sites of 2 and 4 are clustered into the same group by characteristics of hydraulics which come from that of main stream. But their changing pattern of water quality looks like different since the site of 2 is near to dam. The sampling sites of 3, 8, and 9 are individually positioned due to the different tributary.

• International Journal of Computer Science & Network Security
• /
• v.22 no.3
• /
• pp.396-404
• /
• 2022

Online social networks contain a large amount of data that can be converted into valuable and insightful information. Text mining approaches allow exploring large-scale data efficiently. Therefore, this study reviews the recent literature on text mining in online social networks in a way that produces valid and valuable knowledge for further research. The review identifies text mining techniques used in social networking, the data used, tools, and the challenges. Research questions were formulated, then search strategy and selection criteria were defined, followed by the analysis of each paper to extract the data relevant to the research questions. The result shows that the most social media platforms used as a source of the data are Twitter and Facebook. The most common text mining technique were sentiment analysis and topic modeling. Classification and clustering were the most common approaches applied by the studies. The challenges include the need for processing with huge volumes of data, the noise, and the dynamic of the data. The study explores the recent development in text mining approaches in social networking by providing state and general view of work done in this research area.

• Journal of Intelligence and Information Systems
• /
• v.21 no.3
• /
• pp.19-36
• /
• 2015

Social media is an emerging issue in content services and in current business environment. YouTube is the most representative social media service in the world. YouTube is different from other conventional content services in its open user participation and contents creation methods. To promote a content in YouTube, it is important to understand the diffusion phenomena of contents and the network structural characteristics. Most previous studies analyzed impact factors of contents diffusion from the view point of general behavioral factors. Currently some researchers use network structure factors. However, these two approaches have been used separately. However this study tries to analyze the general impact factors on the view count and content based network structures all together. In addition, when building a content based network, this study forms the network structure by analyzing user comments on 22,370 contents of YouTube not based on the individual user based network. From this study, we re-proved statistically the causal relations between view count and not only general factors but also network factors. Moreover by analyzing this integrated research model, we found that these factors affect the view count of YouTube according to the following order; Uploader Followers, Video Age, Betweenness Centrality, Comments, Closeness Centrality, Clustering Coefficient and Rating. However Degree Centrality and Eigenvector Centrality affect the view count negatively. From this research some strategic points for the utilizing of contents diffusion are as followings. First, it is needed to manage general factors such as the number of uploader followers or subscribers, the video age, the number of comments, average rating points, and etc. The impact of average rating points is not so much important as we thought before. However, it is needed to increase the number of uploader followers strategically and sustain the contents in the service as long as possible. Second, we need to pay attention to the impacts of betweenness centrality and closeness centrality among other network factors. Users seems to search the related subject or similar contents after watching a content. It is needed to shorten the distance between other popular contents in the service. Namely, this study showed that it is beneficial for increasing view counts by decreasing the number of search attempts and increasing similarity with many other contents. This is consistent with the result of the clustering coefficient impact analysis. Third, it is important to notice the negative impact of degree centrality and eigenvector centrality on the view count. If the number of connections with other contents is too much increased it means there are many similar contents and eventually it might distribute the view counts. Moreover, too high eigenvector centrality means that there are connections with popular contents around the content, and it might lose the view count because of the impact of the popular contents. It would be better to avoid connections with too powerful popular contents. From this study we analyzed the phenomenon and verified diffusion factors of Youtube contents by using an integrated model consisting of general factors and network structure factors. From the viewpoints of social contribution, this study might provide useful information to music or movie industry or other contents vendors for their effective contents services. This research provides basic schemes that can be applied strategically in online contents marketing. One of the limitations of this study is that this study formed a contents based network for the network structure analysis. It might be an indirect method to see the content network structure. We can use more various methods to establish direct content network. Further researches include more detailed researches like an analysis according to the types of contents or domains or characteristics of the contents or users, and etc.