• Journal of Intelligence and Information Systems
• /
• v.25 no.4
• /
• pp.105-122
• /
• 2019

Dimensionality reduction is one of the methods to handle big data in text mining. For dimensionality reduction, we should consider the density of data, which has a significant influence on the performance of sentence classification. It requires lots of computations for data of higher dimensions. Eventually, it can cause lots of computational cost and overfitting in the model. Thus, the dimension reduction process is necessary to improve the performance of the model. Diverse methods have been proposed from only lessening the noise of data like misspelling or informal text to including semantic and syntactic information. On top of it, the expression and selection of the text features have impacts on the performance of the classifier for sentence classification, which is one of the fields of Natural Language Processing. The common goal of dimension reduction is to find latent space that is representative of raw data from observation space. Existing methods utilize various algorithms for dimensionality reduction, such as feature extraction and feature selection. In addition to these algorithms, word embeddings, learning low-dimensional vector space representations of words, that can capture semantic and syntactic information from data are also utilized. For improving performance, recent studies have suggested methods that the word dictionary is modified according to the positive and negative score of pre-defined words. The basic idea of this study is that similar words have similar vector representations. Once the feature selection algorithm selects the words that are not important, we thought the words that are similar to the selected words also have no impacts on sentence classification. This study proposes two ways to achieve more accurate classification that conduct selective word elimination under specific regulations and construct word embedding based on Word2Vec embedding. To select words having low importance from the text, we use information gain algorithm to measure the importance and cosine similarity to search for similar words. First, we eliminate words that have comparatively low information gain values from the raw text and form word embedding. Second, we select words additionally that are similar to the words that have a low level of information gain values and make word embedding. In the end, these filtered text and word embedding apply to the deep learning models; Convolutional Neural Network and Attention-Based Bidirectional LSTM. This study uses customer reviews on Kindle in Amazon.com, IMDB, and Yelp as datasets, and classify each data using the deep learning models. The reviews got more than five helpful votes, and the ratio of helpful votes was over 70% classified as helpful reviews. Also, Yelp only shows the number of helpful votes. We extracted 100,000 reviews which got more than five helpful votes using a random sampling method among 750,000 reviews. The minimal preprocessing was executed to each dataset, such as removing numbers and special characters from text data. To evaluate the proposed methods, we compared the performances of Word2Vec and GloVe word embeddings, which used all the words. We showed that one of the proposed methods is better than the embeddings with all the words. By removing unimportant words, we can get better performance. However, if we removed too many words, it showed that the performance was lowered. For future research, it is required to consider diverse ways of preprocessing and the in-depth analysis for the co-occurrence of words to measure similarity values among words. Also, we only applied the proposed method with Word2Vec. Other embedding methods such as GloVe, fastText, ELMo can be applied with the proposed methods, and it is possible to identify the possible combinations between word embedding methods and elimination methods.

• Journal of KIISE:Computer Systems and Theory
• /
• v.30 no.5_6
• /
• pp.324-329
• /
• 2003

The public-key cryptosystems such as Diffie-Hellman Key Distribution and Elliptical Curve Cryptosystems are built on the basis of the operations defined in GF(2$^{m}$ ):addition, subtraction, multiplication and multiplicative inversion. It is important that these operations should be computed at high speed in order to implement these cryptosystems efficiently. Among those operations, as being the most time-consuming, multiplicative inversion has become the object of lots of investigation Formant's theorem says $\beta$$^{-1}$ =$\beta$$^{2}$sup m/-2/, where $\beta$$^{-1}$ is the multiplicative inverse of $\beta$$\in$GF(2$^{m}$ ). Therefore, to compute the multiplicative inverse of arbitrary elements of GF(2$^{m}$ ), it is most important to reduce the number of times of multiplication by decomposing 2$^{m}$ -2 efficiently. Among many algorithms relevant to the subject, the algorithm proposed by Itoh and Tsujii[2] has reduced the required number of times of multiplication to O(log m) by using normal basis. Furthermore, a few papers have presented algorithms improving the Itoh and Tsujii's. However they have some demerits such as complicated decomposition processes[3,5]. In this paper, in the case of 2$^{m}$ -2, which is mainly used in practical applications, an efficient algorithm is proposed for computing the multiplicative inverse at high speed by using both the factorization formula x$^3$-y$^3$=(x-y)(x$^2$＋xy＋y$^2$) and normal basis. The number of times of multiplication of the algorithm is smaller than that of the algorithm proposed by Itoh and Tsujii. Also the algorithm decomposes 2$^{m}$ -2 more simply than other proposed algorithms.

• Korean Security Journal
• /
• no.28
• /
• pp.103-130
• /
• 2011

This study is the research of enterprise security for raising the profitability and stability of Korean companies in global business environment and strategic cooperation of business. As the scientific technology gets complicated as day goes by and new competitors appear regardless the border in the modern business environment, the situation happens frequently which the huge company hands over their market to the new one armed with the innovative thinking overnight. To survive such new environment, the answer is the change of paradigm regarding business management method at the new point of view. With the low level of security risk management of Korean companies which stick to old habit, the security management which helps the companies secure profits is not affordable. The global village where the population of 7 billions live in 21st century is facing up to the rapid ecological adaptation. The rapid change of climatic environment creates the hundreds of thousands of sufferers in a moment, and we have been watching the millions of livestock are buried alive due to new contagious disease everyday. Such change encourages the humans in global village to change the basic way of living. The business ecosystem which is the basic root for economic life cannot be an exception. To survive the business environment of 21st century, the security risk management at management level is required and the reporting line of companies should be established newly for raising business competing power through security risk management. The companies should bear in mind that they can be disappeared into our old memories overnight if they are not sensitive to the changing environment. Out of new risks for the modern companies, the field especially Korean companies are dealing easily is the security risk. Not like past, the security risk which's size is much more massive and its propagation velocity is very fast is the one of important business risks which the management should take care. Out of security risks which influence on the modern companies significantly, the brand of companies, protection of their reputation, continuity of production and operation and keeping customer's trust are prior to the others. This study offered the suggestion regarding enterprise security and the strategic cooperation of business to deal with such security risk effectively.

• Korean Security Journal
• /
• no.43
• /
• pp.67-97
• /
• 2015

In Korea, a single emergency number, such as 911 of the USA and 999 of the UK, does not exist. This issue became highly controversial, when the Sewol Ferry Sinking disaster occurred last year. So, the Korean government has planned to adopt a single emergency number, integrating 112 of the Police, 119 of the Fire and Ambulance, 122 of the Korean Coast Guard, and many other emergency numbers. However, the integration plan recently proposed by the Ministry of Public Safety Security seems to be, what is called, a "partial integration model" which repeals the 122 number, but still maintains 112, 119, and 110 respectively. In this context, the study looks into USA's (diverse) 911 operating system, and subsequently tries to draw general features or characteristics. Further, the research attempts to derive policy implication from the general features. If the proposed partial integration model reflects the policy implications, the model can virtually operate like the 911 system -i.e. a single emergency number system - creating inter-operability between responding agencies such as police, fire, and ambulance, even though it is not a perfect integration model. The features drawn are (1) integration of emergency call-taking, (2) functional separation of call-taking and dispatching, (3) integration of physical facilities for call-taking and dispatching, and (4) professional call-takers and dispatchers. Moreover, the policy implications derived from the characteristics are (1) a user-friendly system - fast but accurate responses, (2) integrated responses to accidents, (3) professional call-taking and dispatching & objective and comprehensive risk assessment, and finally (4) active organizational learning in emergency call centers. Considering the policy implications, the following suggestions need to be applied to the current proposed plan: 1. Emergency services' systems should be tightly linked and connected in a systemic way so that they can communicate and exchange intelligence with one another. 2. Public safety answering points (call centers) of each emergency service should share their education and training modules, manuals, etc. Common training and manuals are also needed for inter-operability. 3. Personal management to enable-long term service in public safety answering points (call centers) should be established as one of the ways to promote professionalism.

• Journal of the Korean Vacuum Society
• /
• v.18 no.4
• /
• pp.288-295
• /
• 2009

We report the etch characteristics of GaAs and AlGaAs in the diffusion pump-based capacitively coupled $BCl_3$ plasma. Process variables were chamber pressure ($50{\sim}180$ mTorr), CCP power ($50{\sim}200\;W$) and $BCl_3$ gas flow rate ($2.5{\sim}10$ sccm). Surface profilometry was used for etch rate and surface roughness measurement after etching. Scanning electron microscopy was used to analyze the etched sidewall and surface morphology. Optical emission spectroscopy was used in order to characterize the emission peaks of the $BCl_3$ plasma during etching. We have achieved $0.25{\mu}m$/min of GaAs etch rate with only 5 sccm $BCl_3$ flow rate when the chamber pressure was in the range of 50{\sim}130 mTorr. The etch rates of AlGaAs were a little lower than those of GaAs at the conditions. However, the etch rates of GaAs and AlGaAs decreased significantly when the chamber pressure increased to 180 mTorr. GaAs and AlGaAs were not etched with 50 W CCP power. With $100{\sim}200\;W$ CCP power, etch rates of the materials increased over $0.3{\mu}m$/min. It was found that the etch rates of GaAs and AlGaAs were not always proportional to the increase of CCP power. We also found the interesting result that AlGaAs did not etched at 2.5 sccm $BCl_3$ flow rate at 75 mTorr and 100 W CCP power even though it was etched fast like GaAs with more $BCl_3$ gas flow rates. By contrast, GaAs was etched at ${{\sim}}0.3{\mu}m$/min at the 2.5 sccm $BCl_3$ flow rate condition. A broad molecular peak was noticed in the range of $500{\sim}700\;mm$ wavelength during the $BCl_3$ plasma etching. SEM photos showed that 10 sccm $BCl_3$ plama produced more undercutting on GaAs sidewall than 5 sccm $BCl_3$ plasma.

• Asia pacific journal of information systems
• /
• v.24 no.4
• /
• pp.493-513
• /
• 2014

With the development of related technologies, Location-Based Services (LBS) are growing fast and being used in many ways. Past LBS studies have focused on adoption of LBS because of the fact that LBS users have privacy concerns regarding revealing their location information. Meanwhile, the number of LBS users and revenues from LBS are growing rapidly because users can get some benefits by revealing their location information. Little research has been done on how LBS affects consumers' information search behavior in product purchase. The purpose of this paper is examining the effect of LBS information filtering on buyers' uncertainty and their information search behavior. When consumers purchase a product, they try to reduce uncertainty by searching information. Generally, there are two types of uncertainties - knowledge uncertainty and choice uncertainty. Knowledge uncertainty refers to the lack of information on what kinds of alternatives are available in the market and/or their important attributes. Therefore, consumers having knowledge uncertainty will have difficulties in identifying what alternatives exist in the market to fulfil their needs. Choice uncertainty refers to the lack of information about consumers' own preferences and which alternative will fit in their needs. Therefore, consumers with choice uncertainty have difficulties selecting best product among available alternatives.. According to economics of information theory, consumers narrow the scope of information search when knowledge uncertainty is high. It is because consumers' information search cost is high when their knowledge uncertainty is high. If people do not know available alternatives and their attributes, it takes time and cognitive efforts for them to acquire information about available alternatives. Therefore, they will reduce search breadth. For people with high knowledge uncertainty, the information about products and their attributes is new and of high value for them. Therefore, they will conduct searches more in-depth because they have incentive to acquire more information. When people have high choice uncertainty, people tend to search information about more alternatives. It is because increased search breadth will improve their chances to find better alternative for them. On the other hand, since human's cognitive capacity is limited, the increased search breadth (more alternatives) will reduce the depth of information search for each alternative. Consumers with high choice uncertainty will spend less time and effort for each alternative because considering more alternatives will increase their utility. LBS provides users with the capability to screen alternatives based on the distance from them, which reduces information search costs. Therefore, it is expected that LBS will help users consider more alternatives even when they have high knowledge uncertainty. LBS provides distance information, which helps users choose alternatives appropriate for them. Therefore, users will perceive lower choice uncertainty when they use LBS. In order to test the hypotheses, we selected 80 students and assigned them to one of the two experiment groups. One group was asked to use LBS to search surrounding restaurants and the other group was asked to not use LBS to search nearby restaurants. The experimental tasks and measures items were validated in a pilot experiment. The final measurement items are shown in Appendix A. Each subject was asked to read one of the two scenarios - with or without LBS - and use a smartphone application to pick a restaurant. All behaviors on smartphone were recorded using a recording application. Search breadth was measured by the number of restaurants clicked by each subject. Search depths was measured by two metrics - the average number of sub-level pages each subject visited and the average time spent on each restaurant. The hypotheses were tested using SPSS and PLS. The results show that knowledge uncertainty reduces search breadth (H1a). However, there was no significant correlation between knowledge uncertainty and search depth (H1b). Choice uncertainty significantly reduces search depth (H2b), but no significant relationship was found between choice uncertainty and search breadth (H2a). LBS information filtering significantly reduces the buyers' choice uncertainty (H4) and reduces the negative relationship between knowledge uncertainty and search breadth (H3). This research provides some important implications for service providers. Service providers should use different strategies based on their service properties. For those service providers who are not well-known to consumers (high knowledge uncertainty) should encourage their customers to use LBS. This is because LBS would increase buyers' consideration sets when the knowledge uncertainty is high. Therefore, less known services have chances to be included in consumers' consideration sets with LBS. On the other hand, LBS information filtering decrease choice uncertainty and the near service providers are more likely to be selected than without LBS. Hence, service providers should analyze geographically approximate competitors' strength and try to reduce the gap so that they can have chances to be included in the consideration set.

• Journal of the Korean Society for Nondestructive Testing
• /
• v.26 no.4
• /
• pp.211-219
• /
• 2006

The UET(ultrasound excited thermography) for the ,eat-time diagnostics of the object employs an infrared camera to image defects of the surface and subsurface which are locally heated using high-frequency putted ultrasonic excitation. The dissipation of high-power ultrasonic energy around the feces of the defects causes an increase In temperature. The defect's image appears as a hot spot (bright IR source) within a dark background field. The UET for nondestructive diagnostic and evaluation is based on the image analysis of the hot spot as a local response to ultrasonic excited heat deposition. In this paper the applicability of VET for fast imaging of defect is described. The ultrasonic energy is injected into the sample through a transducer in the vertical and horizontal directions respectively. The voltage applied to the transducer is measured by digital oscilloscope, and the waveform are compared. Measurements were performed on four kinds of materials: SUS fatigue crack specimen(thickness 14mm), PCB plate(1.8 mm), CFRP plate(3 mm) and Inconel 600 plate (1 mm). A high power ultrasonic energy with pulse durations of 250ms Is injected into the samples in the horizontal and vertical directions respectively The obtained experimental result reveals that the dissipation loss of the ultrasonic energy In the vertical injection is less than that in the horizontal direction. In the cafe or PCB, CFRP, the size of hot spot in the vortical injection if larger than that in horizontal direction. Duration time of the hot spot in the vertical direction is three times as long as that in the horizontal direction. In the case of Inconel 600 plate and SUS sample, the hot spot in the horizontal injection was detected faster than that in the vertical direction

• The Sea:JOURNAL OF THE KOREAN SOCIETY OF OCEANOGRAPHY
• /
• v.8 no.4
• /
• pp.401-410
• /
• 2003

Three beaches of the Seogwang-ri coast in the western part of Wu Island, Jeju-do, are solely composed of rhodoliths (red algal nodules). The beach sediments are coarse sand to granule in size and they show the banded distribution according to size. Commonly the larger pebble-sized rhodoliths are concentrated near the rocky coast, resulting from the transportation of the nodules from shallow marine environments by intermittent typhoons. Based on the internal texture of the rhodoliths, it appears that crustose red algae, Lithophyllum sp., is the main contributor for the formation of the rhodolith. The coarse sand to granule-sized grains show that they started to grow from the nucleus as rhodoliths, but the surface was severely eroded by waves. However, the pebble to cobble-sized grains exhibit the complete growth pattern of rhodoliths and sometimes contain other calcareous skeletons. It is common that encrusting red algae are intergrown with encrusting bryozoan. The surface morphology of rhodolith tends to change from the concentric to domal shape towards the outer part. This suggests that the rhodolith grew to a certain stage by rolling, but it grew in more quiet condition without rolling as it became larger. Aragonite and calcite cements can be found in the pores within rhodoliths (conceptacle, intraskeletal pore in bryozoan, and boring), and this means that shallow marine cementation has occurred during their growth. Growth of numerous rhodoliths in shallow marine environment near the Seogwang-ri coast indicates that this area has suitable oceanographic conditions for their growth such as warm water temperature (about 19$^{\circ}C$ in average) and clear water condition due to the lack of terrestrial input of volcanoclastic sediments. Fast tidal current and high wave energy in the shallow water setting can provide suitable conditions enough for their rolling and growth. Typhoons passing this area every summer also influence on the growth of rhodoliths.

• The Korean Journal of Food And Nutrition
• /
• v.24 no.1
• /
• pp.124-131
• /
• 2011

We analyzed content and peptides in order to investigate the productivity from Cheonggookjang(fast-fermented soybean paste), fermenting it for 180 hours at $40^{\circ}C$. Results showed that pH was 7.07 at the start and became 7.41 in 24 hours, it eventually increased to 8.63 after 180 hours. Acidity was 0.2 in 12 hours, 0.5 in 12 hours, and then remained on 0.1 thereafter. Total sugar was 1.54 mg/$m\ell$ at the start, but it gradually decreased to 0.76 mg/$m\ell$ after the lapse of 48 hours, and 1.0 mg/$m\ell$ in 120 hours, and finally 0.8 mg/$m\ell$ in 180 hours. Reducing sugar was 0.14 mg/$m\ell$ at the start, and 0.88 mg/$m\ell$ after the lapse of 24 hours, 0.64 mg/$m\ell$ in 48 hours, 0.26 mg/$m\ell$ in 72 hours, and showed no definite change untill 180 hours. The amount of free amino acid was $0.19\;{\mu}M/\ell$ at the start, and $4.88\;{\mu}M/\ell$ after the lapse of 72 hours, $4.5\;{\mu}M/\ell$ in 120 hours, and then it rapidly decreased to $0.23\;{\mu}M/\ell$ after180 hours. Absorbance of soluble protein and peptide at 280 nm was 12.4 in 48 hours, 31.12 in 120 hours, and 31.12 in 180 hours. HPLC revealed that in the fermentation process, large molecular proteins are hydrolyzed into small peptides and amino acids, and after the lapse of 48 hours the pattern became almost the same. The protease activity of Cheonggookjang was 0.011 unit/$m\ell$ after the lapse of 36 hours and then it decreased. The result shows as Cheonggookjang started its deamination of amino acid in 100 hours, it is desirable to produce peptide within 100 hours of its fermentation.

• Journal of Nutrition and Health
• /
• v.52 no.6
• /
• pp.559-568
• /
• 2019

Purpose: This study examined the prevalence of food allergies and allergenic factors in a selected sample of children living in Seoul, Korea, along with their dietary habits, environmental factors, and diseases as risk factors for food allergy. The results of this study will provide basic data for addressing food allergies. Methods: We selected 3,004 pre-school and school-age children, aged 0 ~ 12, in the 25 districts of Seoul as the study sample. Structured self-report questionnaires were administered over a two-month period in July-August 2018, and the children's parents recorded the answers on their children's behalf. The research tools in this study included the Korean version of the questionnaire from the International Study of Asthma and Allergies in Childhood (ISAAC). Results: The physician-diagnosed prevalence rate of food allergies was 14.2%, while 20.4% of the children experienced allergic symptoms at least once and 17.4% reported symptoms within the previous 12 months. The children's symptoms included skin problems (88.1%), gastrointestinal issues (19.2%), oral issues (16.7%), respiratory issues (12.7%), and systemic issues (1.3%). The causes of allergies included eggs, peaches, milk, peanuts, and shrimps. The factors influencing the experience of food allergies were the consumption of cereal (aOR, 1.52; 95% CI, 1.09 ~ 2.10; p = 0.013), potatoes (aOR, 1.88; 95% CI, 1.33 ~ 2.65; p < 0.001), and fast food (aOR, 1.73; 95% CI, 1.10 ~ 2.72; p = 0.017). Having food allergy symptoms was associated with a higher risk of experiencing asthma (aOR, 4.22 95% CI, 3.10 ~ 5.76; p < 0.001), allergic rhinitis (aOR, 2.53; 95% CI, 2.03 ~ 3.15; p < 0.001), and atopic dermatitis symptoms (aOR, 3.56; 95% CI, 2.88 ~ 4.40; p < 0.001). Conclusion: Episodes of food allergies warrant examining regular food consumption and placing dietary restrictions through early diagnosis as these episodes may imply the presence of other allergies. Our findings offer basic insights into the patterns, prevalence and symptoms of children's food allergies in Seoul, and our findings will contribute to identifying effective interventions for food allergies.