• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.1
• /
• pp.305-326
• /
• 2019

Discriminative correlation filter (DCF) based tracking algorithms have recently shown impressive performance on benchmark datasets. However, amount of recent researches are vulnerable to heavy occlusions, irregular deformations and so on. In this paper, we intend to solve these problems and handle the contradiction between accuracy and real-time in the framework of tracking-by-detection. Firstly, we propose an innovative strategy to combine the template and color-based models instead of a simple linear superposition and rely on the strengths of both to promote the accuracy. Secondly, to enhance the discriminative power of the learned template model, the spatial regularization is introduced in the learning stage to penalize the objective boundary information corresponding to features in the background. Thirdly, we utilize a discriminative multi-scale estimate method to solve the problem of scale variations. Finally, we research strategies to limit the computational complexity of our tracker. Abundant experiments demonstrate that our tracker performs superiorly against several advanced algorithms on both the OTB2013 and OTB2015 datasets while maintaining the high frame rates.

• Language and Information
• /
• v.14 no.2
• /
• pp.159-181
• /
• 2010

This paper presents a hybrid approach to the sentiment analysis of online texts. The sentiment of a text refers to the feelings that the author of a text has towards a certain topic. Many existing approaches employ either a pattern-based approach or a machine learning based approach. The former shows relatively high precision in classifying the sentiments, but suffers from the data sparseness problem, i.e. the lack of patterns. The latter approach shows relatively lower precision, but 100% recall. The approach presented in the current work adopts the merits of both approaches. It combines the pattern-based approach with the machine learning based approach, so that the relatively high precision and high recall can be maintained. Our experiment shows that the hybrid approach improves the F-measure score for more than 50% in comparison with the pattern-based approach and for around 1% comparing with the machine learning based approach. The numerical improvement from the machine learning based approach might not seem to be quite encouraging, but the fact that in the current approach not only the sentiment or the polarity information of sentences but also the additional information such as target of sentiments can be classified makes the current approach promising.

• Journal of the Korean Data and Information Science Society
• /
• v.21 no.6
• /
• pp.1263-1270
• /
• 2010

Sixty patients were divided into three groups. Each group of twenty persons had fed on different diet foods over 5 weeks. Cholesterol had been measured repeatedly five times at an interval of a week during 5 weeks. It resulted from mixed model analysis of repeated measurements data that homogeneous toeplitz covariance matrix pattern was selected as the optimal covariance pattern. The correlations between measurements of different times for the covariance matrix are somewhat highly correlated as 0.64-0.78. Based upon the homogeneous toeplitz covariance pattern model, the time effect was found to be highly significant, but the treatment effect and treatment-time interaction effect were found to be insignificant.

• Journal of the Korean Data and Information Science Society
• /
• v.14 no.4
• /
• pp.725-736
• /
• 2003

To provide the information needed to make a decision, it is important to know the relationship or pattern between variables in database. Grouping objects which have similar characteristics of pattern is called as cluster analysis, one of data mining techniques. In this study, it is compared with several partitioning clustering algorithms, based on the statistical distance or total variance in each cluster.

• Journal of the Korean Society for information Management
• /
• v.14 no.1
• /
• pp.27-45
• /
• 1997

This study is intended to investigate and analyze overall information use pattern of researchers and scholars in the field of science and social science, and compare the information use habits between two groups. 300 questionnaires were sent out to 7 major libraries or information centers in the field, and 217 questionnaires were returned. Five hypotheses were tested based on the assumption that there would exist some differences between two groups in their information use pattern. Some commonalities a d differences were revealed as the findings of the study.

• Proceedings of the Korea Association of Information Systems Conference
• /
• 2005.05a
• /
• pp.79-92
• /
• 2005

To query the vast amount of web pages which are available i]l the Internet, it is necessary to extract the encoded information in the web pages for converting it into structured data (e.g. relational data for SQL) or semistructured data (e.g. XML data for XQuery), In this paper, we propose a new web information extraction system, PIES, to convert web information into XML documents. PIES is based on a user-specified target schema and HTML tag pattern descriptions. The web information is extracted by the pattern descriptions and validated by the target schema. We designed a new language to describe extraction rules, and a new regular expression to describe HTML tag patterns. We implemented PIES and applied it to the US patent web site to evaluate its correctness. It successfully extracted more than thousands of US patent data and converted them into XML documents.

• Communications for Statistical Applications and Methods
• /
• v.8 no.1
• /
• pp.271-279
• /
• 2001

Sequential pattern discovery is one of main interests in web usage mining. the technique of sequential pattern discovery attempts to find inter-session patterns such that the presence of a set of items is followed by another item in a time-ordered set of server sessions. In this paper, a tree-based sequential pattern finding method is proposed in order to discover navigation patterns in server sessions. At each learning process, the suggested method learns about the navigation patterns per server session and summarized into the modified Rymon's tree.

• Journal of Korean Institute of Industrial Engineers
• /
• v.28 no.1
• /
• pp.112-127
• /
• 2002

A useful pattern is a pattern that contributes much to learning. For a classification problem those patterns near the class boundary surfaces carry more information to the classifier. For a regression problem the ones near the estimated surface carry more information. In both cases, the usefulness is defined only for those patterns either without error or with negligible error. Using only the useful patterns gives several benefits. First, computational complexity in memory and time for learning is decreased. Second, overfitting is avoided even when the learner is over-sized. Third, learning results in more stable learners. In this paper, we propose a pattern 'utility index' that measures the utility of an individual pattern. The utility index is based on the bias and variance of a pattern trained by a network ensemble. In classification, the pattern with a low bias and a high variance gets a high score. In regression, on the other hand, the one with a low bias and a low variance gets a high score. Based on the distribution of the utility index, the original training set is divided into a high-score group and a low-score group. Only the high-score group is then used for training. The proposed method is tested on synthetic and real-world benchmark datasets. The proposed approach gives a better or at least similar performance.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2000.05a
• /
• pp.296-300
• /
• 2000

In previous papers, it proved relevant to using together with EFCI and ER switch for effective ABR traffic managements. It also applied to EFCI and ER switch algorithm, that consider ABR traffic as poisson pattern. However, in recently network environment, it has been proved about traffic pattern, that is similar to self-similar pattern than poisson pattern. In this paper, we will compare previous poisson pattern with self-similar pattern under ATM network.

• Structural Monitoring and Maintenance
• /
• v.6 no.1
• /
• pp.67-84
• /
• 2019

Digital Image Correlation technique (DIC) is a non-contact optical method for rapid structural health monitoring of critical infrastructure. An innovative approach to DIC is presented using QR (Quick Response) code based random speckle pattern. Reinforced Cement Concrete (RCC) beams of size $1800mm{\times}150mm{\times}200mm$ are tested in flexure. DIC is used to extract Moment (M) - Curvature (${\kappa}$) relationships using random speckle patterns and QR code based random speckle patterns. The QR code based random speckle pattern is evaluated for 2D DIC measurements and the QR code speckle pattern performs satisfactorily in comparison with random speckle pattern when considered in the context of serving a dual purpose. Characteristics of QR code based random speckle pattern are quantified and its applicability to DIC is explored. The ultimate moment-curvature values computed from the QR code based random speckled pattern are found to be in good agreement with conventional measurements. QR code encrypts the structural information which enables integration with building information modelling (BIM).