• Journal of Internet Computing and Services
• /
• v.12 no.3
• /
• pp.131-137
• /
• 2011

GPUs were originally designed for graphic processing, and GPGPUs are general-purpose GPUs for numerical computation with high performance and low electric power. In this paper, we implemented the parallel LU factorization program for GPGPUs. In CUDA, which is computational environment for Nvidia GPGPUs, domains are divided into blocks, and multi-threads compute each sub-blocks Simultaneously. In LU factorization program, computation order should be artificially decided due to the data dependence. To resolve the data dependancy, we suggested a parallel LU program for GPGPUs, and also explained parallel reduction algorithm for partial pivoting of LU factorization. We finally present performance analysis to show efficiency of the parallel LU factorization program based on multi-threads on GPGPUs.

• Journal of KIISE:Software and Applications
• /
• v.36 no.5
• /
• pp.347-352
• /
• 2009

Nonnegative matrix factorization (NMF) is a popular method for multivariate analysis of nonnegative data, the goal of which is decompose a data matrix into a product of two factor matrices with all entries in factor matrices restricted to be nonnegative. NMF was shown to be useful in a task of clustering (especially document clustering). In this paper we present an algorithm for orthogonal nonnegative matrix factorization, where an orthogonality constraint is imposed on the nonnegative decomposition of a term-document matrix. We develop multiplicative updates directly from true gradient on Stiefel manifold, whereas existing algorithms consider additive orthogonality constraints. Experiments on several different document data sets show our orthogonal NMF algorithms perform better in a task of clustering, compared to the standard NMF and an existing orthogonal NMF.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.9
• /
• pp.3421-3443
• /
• 2021

A data hiding method based on integer factorization via e-books in the EPUB format with XHTML and CSS files for copyright protection is proposed. Firstly, a fixed number m of leading bits in a message are transformed into an integer which is then factorized to yield k results. One of the k factorizations is chosen according to the decimal value of a number n of the subsequent message bits with n being decided as the binary logarithm of k. Next, the chosen factorization, denoted as a × b, is utilized to create a combined use of the

and elements in the XHTML files to embed the m + n message bits by including into the two elements a class selector named according to the value of a as well as a text segment with b characters. The class selector is created by the use of a CSS pseudo-element. The resulting web pages are of no visual difference from the original, achieving a steganographic effect. The security of the embedded message is also considered by randomizing the message bits before they are embedded. Good experimental results and comparisons with exiting methods show the feasibility of the proposed method for copyright protection of e-books.

• Journal of Digital Contents Society
• /
• v.18 no.4
• /
• pp.707-712
• /
• 2017

As the amount of data increases exponentially, the recommender system is attracting interest in various industries such as movies, books, and music, and is being studied. The recommendation system aims to propose an appropriate item to the user based on the user's past preference and click stream. Typical examples include Netflix's movie recommendation system and Amazon's book recommendation system. Previous studies can be categorized into three types: collaborative filtering, content-based recommendation, and hybrid recommendation. However, existing recommendation systems have disadvantages such as sparsity, cold start, and scalability problems. To improve these shortcomings and to develop a more accurate recommendation system, we have designed a recommendation system as a factorization machine using actual online product purchase data.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.17 no.3
• /
• pp.721-739
• /
• 2023

Rating prediction is an important issue in recommender systems, and its accuracy affects the experience of the user and the revenue of the company. Traditional recommender systems use Factorization Machinesfor rating predictions and each feature is selected with the same weight. Thus, there are problems with inaccurate ratings and limited data representation. This study proposes a deep recommendation model based on self-attention Factorization (SAFMR) to solve these problems. This model uses Convolutional Neural Networks to extract features from user and item reviews. The obtained features are fed into self-attention mechanism Factorization Machines, where the self-attention network automatically learns the dependencies of the features and distinguishes the weights of the different features, thereby reducing the prediction error. The model was experimentally evaluated using six classes of dataset. We compared MSE, NDCG and time for several real datasets. The experiment demonstrated that the SAFMR model achieved excellent rating prediction results and recommendation correlations, thereby verifying the effectiveness of the model.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.47 no.11
• /
• pp.54-64
• /
• 2010

Reed-Solomon (RS) codes are the most widely used error correcting codes in digital communications and data storage. Recently, Sudan found algorithm of list decoder for RS codes. List decoder has larger decoding radius than conventional hard-decision decoding algorithms and return more than one candidate polynomial. But, the algorithm includes interpolation and factorization step that demand massive computations. In this paper, an efficient architecture and processing schedule are proposed. The architecture consists of R-MAC, memories, and control unit. The R-MAC computes both of RC and PU steps that are main part of the factorization algorithm. The proposed architecture can achieve higher hardware utilization efficiency (HUE) and throughput by using efficient processing schedule and memory architecture. Also, the architecture can be designed flexibly with scalability for various applications. We design and synthesize our architecture using Dongbu-Anam $0.18{\mu}m$ standard cell library and the maximum clock frequency is 330MHz.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.5
• /
• pp.1163-1185
• /
• 2024

With the sharp increase in the volume of literature data, researchers must spend considerable time and energy locating desired papers. A paper recommendation is the means necessary to solve this problem. Unfortunately, the large amount of data combined with sparsity makes personalizing papers challenging. Traditional matrix decomposition models have cold-start issues. Most overlook the importance of information and fail to consider the introduction of noise when using side information, resulting in unsatisfactory recommendations. This study proposes a paper recommendation method (PR-SLSMF) using document-level representation learning with citation-informed transformers (SPECTER) and low-rank and sparse matrix factorization; it uses SPECTER to learn paper content representation. The model calculates the similarity between papers and constructs a weighted heterogeneous information network (HIN), including citation and content similarity information. This method combines the LSMF method with HIN, effectively alleviating data sparsity and cold-start issues and avoiding topic drift. We validated the effectiveness of this method on two real datasets and the necessity of adding side information.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.393-396
• /
• 2011

In this paper, we propose a constrained alternating least squares nonnegative matrix factorization algorithm (cALSNMF) to detect active brain regions from single subject's task-related fMRI data. In cALSNMF, we define a new cost function which considers the uncorrelation and noisy problems of fMRI data by adding decorrelation and smoothing constraints in original Euclidean distance cost function. We also generate a novel training procedure by modifying the update rules and combining with optimal brain surgeon (OBS) algorithm. The experimental results on visuomotor task fMRI data show that our cALSNMF fits fMRI data better than original ALSNMF in detecting task-related brain activation from single subject's fMRI data.

• Journal of KIISE:Computing Practices and Letters
• /
• v.14 no.5
• /
• pp.497-501
• /
• 2008

In this paper we present a method for continuous EEG classification, where we employ nonnegative tensor factorization (NTF) to determine discriminative spectral features and use the Viterbi algorithm to continuously classily multiple mental tasks. This is an extension of our previous work on the use of nonnegative matrix factorization (NMF) for EEG classification. Numerical experiments with two data sets in BCI competition, confirm the useful behavior of the method for continuous EEG classification.

• The Journal of the Acoustical Society of Korea
• /
• v.31 no.5
• /
• pp.317-331
• /
• 2012

In this paper, we develop a multichannel blind source separation algorithm based on a beamspace transform and the multichannel non-negative matrix factorization (NMF) method. The NMF algorithm is a famous algorithm which is used to solve the source separation problems. In this paper, we consider a beamspace-time-frequency domain data model for multichannel NMF method, and enhance the conventional method using a beamspace transform. Our decomposition algorithm is applied to audio source separation, using a dataset from the international Signal Separation Evaluation Campaign 2010 (SiSEC 2010) for evaluation.