• Journal of the Institute of Convergence Signal Processing
• /
• v.2 no.3
• /
• pp.16-23
• /
• 2001

In this paper, we propose the pseudo n-gram language models for speech recognition with middle size vocabulary compared to large vocabulary speech recognition using the statistical n-gram language models. The proposed method is that it is very simple method, which has the standard structure of ARPA and set the word probability arbitrary. The first, the 1-gram sets the word occurrence probability 1 (log likelihood is 0.0). The second, the 2-gram also sets the word occurrence probability 1, which can only connect the word start symbol and WORD, WORD and the word end symbol . Finally, the 3-gram also sets the ward occurrence probability 1, which can only connect the word start symbol , WORD and the word end symbol . To verify the effectiveness of the proposed method, the word recognition experiments are carried out. The preliminary experimental results (off-line) show that the word accuracy has average 97.7% for 452 words uttered by 3 male speakers. The on-line word recognition results show that the word accuracy has average 92.5% for 20 words uttered by 20 male speakers about stock name of 1,500 words. Through experiments, we have verified the effectiveness of the pseudo n-gram language modes for speech recognition.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.13 no.4
• /
• pp.805-812
• /
• 2009

In many intrusion detection applications, n-gram approach has been widely applied. However, n-gram approach has shown a few problems including unscalability and double counting of features. To address those problems, we applied n-gram augmented Naive Bayes with k-truncated suffix tree (k-TST) storage mechanism directly to classify intrusive sequences and compared performance with those of Naive Bayes and Support Vector Machines (SVM) with n-gram features by the experiments on host-based intrusion detection benchmark data sets. Experimental results on the University of New Mexico (UNM) benchmark data sets show that the n-gram augmented method, which solves the problem of independence violation that happens when n-gram features are directly applied to Naive Bayes (i.e. Naive Bayes with n-gram features), yields intrusion detectors with higher accuracy than those from Naive Bayes with n-gram features and shows comparable accuracy to those from SVM with n-gram features. For the scalable and efficient counting of n-gram features, we use k-truncated suffix tree mechanism for storing n-gram features. With the k-truncated suffix tree storage mechanism, we tested the performance of the classifiers up to 20-gram, which illustrates the scalability and accuracy of n-gram augmented Naive Bayes with k-truncated suffix tree storage mechanism.

• Korean Journal of Microbiology
• /
• v.47 no.1
• /
• pp.50-55
• /
• 2011

A simple and rapid method was developed for monitoring of Gram-positive bacteria in the wastewater treatment system. Culture suspensions of 4 Gram-positive and 4 Gram-negative strains were filtrated and stained with a polyethersulfone membrane filter and Toluidine Blue-O. To establish quantitative color image analysis, the intensity value of RGB (red-green-blue) color of a scanned filter image was analyzed with a photographic program. Red and green color values of Gram-positive bacteria were higher than those of Gram-negative bacteria. This method was applied to the activated sludge mixed with the Gram-positive bacteria. Although evaluation was difficult due to the irregular size and shape of flocs, the population of Gram-positive bacteria in the activated sludge could be monitored with floc dispersion technique. The more amounts of Gram-positive bacteria in the activated sludge led to the increase of red and green color values. This method provides a rapid and quantitative measurement of Gram-positive bacteria within the wastewater treatment systems.

• The Korea Journal of Herbology
• /
• v.37 no.1
• /
• pp.11-18
• /
• 2022

Objectives : Green gram (mung bean) has a cold nature and has been known to detoxify against various side effects that belong to hot in nature. In particular, since it has the effect of detoxifying fever and detoxification to treat swelling, it was also used externally to treat febrile dermatological diseases such as erysipelas and rubella. This study was designed to determine whether green gram exhibits anti-inflammatory effects on contact dermatitis in mice. Methods : We investigated the effects of green gram extract (70% ethanol extract) on skin lesion, skin thickness and weights, melanin and erythema index and spleen body weight ratio in mice with contact dermatitis induced by repeated application of 1-Fluoro-2,4-dinitrobenzene. Results : Topical application of green gram extract ameliorates skin lesions of contact dermatitis such as scale and roughness induces by 1-Fluoro-2,4-dinitrobenzene. green gram extract also suppressed enlargement of skin thicknesses and weights significantly. In addition, green gram extract treatment also lowered erythema index significantly compared to those in the control group. In the histopathological observation, green gram extract prevented epidermal hyperplasia and hyperkeratosis in inflamed tissues. Finally, green gram extract did not affect changes in body weights and the spleen body weight ratio, unlike dexamethasone, which significantly prevented body weight gain and lowered the spleen body weight ratio. Conclusions : These results imply that green gram, which is known to have a detoxifying effect in Korean medicine, can be used in the treatment of contact dermatitis.

• Annual Conference on Human and Language Technology
• /
• 2019.10a
• /
• pp.496-499
• /
• 2019

스마트폰의 쿼티 자판 소프트 키보드의 버튼과 버튼 사이 좁은 간격으로 인해 사용자가 의도치 않은 간섭 오타가 발생하는 것에 주목하였다. 그리고 오타 교정의 성능은 사용자의 관점에서 얼마나 잘 오타를 교정하느냐도 중요한 부분이지만, 또한 오타가 아닌 어절을 그대로 유지하는 것이 더 중요하게 판단될 수 있다. 왜냐하면 현실적으로 오타인 어절 보다 오타가 아닌 어절이 거의 대부분을 차지하기 때문이다. 따라서 해당 관점에서 교정 방법을 바라보고 연구할 필요가 있다. 이에 맞춰 본 논문에서는 대용량 한국어 말뭉치 데이터를 가지고 확률에 기반한 한국어 간섭 오타 수정 방법에 대해 제안한다. 제안하는 방법은 목표 어절의 좌우 어절 N-gram과 어절 내 좌우 음절 N-gram 정보를 바탕으로 발생할 수 있는 간섭 오타 교정 후보들 중 가운데서 가장 적합한 후보 어절을 선택하는 방법이다.

• Bulletin of the Korean Mathematical Society
• /
• v.41 no.4
• /
• pp.739-777
• /
• 2004

An analogue of the Gram's inequality for n-inner product spaces is given. Further, a number of inequalities involving Gram's determinant are stated and proved in terms of n-inner products.

• Honam Mathematical Journal
• /
• v.45 no.3
• /
• pp.418-432
• /
• 2023

In this study, we consider the Hankel and Gram matrices which are defined by the elements of special number sequences. Firstly, the eigenvalues, determinant, and norms of the Hankel matrix defined by special number sequences are obtained. Afterwards, using the relationship between the Gram and Hankel matrices, the eigenvalues, determinants, and norms of the Gram matrices defined by number sequences are given.

• Applied Biological Chemistry
• /
• v.17 no.2
• /
• pp.117-124
• /
• 1974

The 23 representative bacteria were studied for the glucosamine contents which represent the glycolipid content of the cell wall. The distribution of glycolipid in various bacteria was examined and the relationship between the glycolipid contents and the Gram stain was elucidated. The results were as follows: 1. The contents of glucosamine in the glycolipid of Gram negative and variable bacterial cell wall were large ranging from the least 0.04 ${\mu}g$ of Proteus vulgaris to the largest 2.48 ${\mu}g$ of Aerobacter aerogenes. The Gram positive bacteria and only those Gram positive among Bacilli contained less than 0.02 ${\mu}g$ of glucosamine contents. The least glucosamine containing Gram positive bacteria were Corynebacterium sepedonicum and Staphylococcus aureus. It could generally be said that the Gram negative and variable bacteria contain the higher content of glucosamine in the cell wall than the positives. 2. The bacteria were better stained by the Gram solution after the extraction of glycolipid from the cell wall than those without extraction. 3. The four infrared spectra of glycolipids obtained from Aerobacter aerogenes, Bacillus circulans, Pseudomonas fluorescens, and Salmonella typhirurium showed all the similar characteristics. All showed the existence of groups; OH, C-O, C-O-C, $CH_2+CH_3$, amide band, fatty acid ester band and ester carbonyl bond.

• Journal of KIISE:Software and Applications
• /
• v.35 no.12
• /
• pp.774-779
• /
• 2008

In this work, we describe comments classification system using topic signature. Topic signature is widely used for selecting feature in document classification and summarization. Comments are short and have so many word spacing errors, special characters. We firstly convert comments into 7-gram. We consider the 7-gram as sentence. We convert the 7-gram into 3-gram. We consider the 3-gram as word. We select key feature using topic signature and classify new inputs by the Naive Bayesian method. From the result of experiments, we can see that the proposed method is outstanding over the previous methods.

• 한국해양학회지
• /
• v.24 no.3
• /
• pp.148-156
• /
• 1989

This investigation was performed in eutrophic lake within the framework of a series of studies to evaluate the significance of gram reaction for both bacterioplankton and attached bacteria in the dynamics of organic materials at various aquatic ecosystems. In Lake Kasumigaura as a representative of the highly eutrophic freshwater environments, the gram-stain characteristics of the bacterial community changed with the influx of pulses of phytoplankton, as those in the meso trophic environments. The predominency of the gram-negative forms in the bacterial community was about 57% for bacterioplankton and about 53% for attached bacteria. The statistical analysis of the difference of these two distributions showed that these communites were different. Both gram-negative and gram-positive bacteria attached to particles were shown to effect the formation and degradation of particulate organic matter. Gram-negative bacteria plankton participate exclusively in the dynamics of dissolved organic matter.