• The KIPS Transactions:PartB
• /
• v.15B no.5
• /
• pp.457-464
• /
• 2008

In the real-world operational environment, most of text classification systems have the problems of insufficient training documents and no prior knowledge of feature space. In this regard, $Na{\ddot{i}ve$ Bayes is known to be an appropriate algorithm of operational text classification since the classification model can be evolved easily by incrementally updating its pre-learned classification model and feature space. This paper proposes the improving technique of $Na{\ddot{i}ve$ Bayes classifier through feature weighting strategy. The basic idea is that parameter estimation of $Na{\ddot{i}ve$ Bayes considers the degree of feature importance as well as feature distribution. We can develop a more accurate classification model by incorporating feature weights into Naive Bayes learning algorithm, not performing a learning process with a reduced feature set. In addition, we have extended a conventional feature update algorithm for incremental feature weighting in a dynamic operational environment. To evaluate the proposed method, we perform the experiments using the various document collections, and show that the traditional $Na{\ddot{i}ve$ Bayes classifier can be significantly improved by the proposed technique.

• Journal of the Korean Institute of Gas
• /
• v.19 no.2
• /
• pp.45-53
• /
• 2015

In order to improve the performance of the visual inspection process, in addition to existing automatic visual inspection machine and human inspectors have developed a new process configuration using a Naive Bayes classifier. By applying the classifier, defect leakage and human inspector's work amount could be improved at the same time. New classification method called AMPB was applied instead of conventional methods based on MAP classification. By experimental results using the filter product for camera modules, it was confirmed that it is possible to configure the process at the level of leakage ratio 1.14% and human inspector's work amount ratio 75.5%. It is significant that the result can be applied in such a wide range as gas leak detection which is the collaboration process between inspection machine and human inspector's

• Journal of KIISE:Computing Practices and Letters
• /
• v.15 no.5
• /
• pp.345-349
• /
• 2009

Social bookmarking systems are a typical web 2.0 service based on folksonomy, providing the platform for storing and sharing bookmarking information. Spammers in social bookmarking systems denote the users who abuse the system for their own interests in an improper way. They can make the entire resources in social bookmarking systems useless by posting lots of wrong information. Hence, it is important to detect spammers as early as possible and protect social bookmarking systems from their attack. In this paper, we applied a diverse set of machine learning approaches, i.e., decision tables, decision trees (ID3), $na{\ddot{i}}ve$ Bayes classifiers, TAN (tree-augment $na{\ddot{i}}ve$ Bayes) classifiers, and artificial neural networks to this task. In our experiments, $na{\ddot{i}}ve$ Bayes classifiers performed significantly better than other methods with respect to the AUC (area under the ROC curve) score as veil as the model building time. Plausible explanations for this result are as follows. First, $na{\ddot{i}}ve$> Bayes classifiers art known to usually perform better than decision trees in terms of the AUC score. Second, the spammer detection problem in our experiments is likely to be linearly separable.

• Journal of the Korean Society for information Management
• /
• v.22 no.2 s.56
• /
• pp.125-145
• /
• 2005

This paper presents a semantic vector space retrieval model incorporating a word sense disambiguation algorithm in an attempt to improve retrieval effectiveness. Nine Korean homonyms are selected for the sense disambiguation and retrieval experiments. The total of approximately 120,000 news articles comprise the raw test collection and 18 queries including homonyms as query words are used for the retrieval experiments. A Naive Bayes classifier and EM algorithm representing supervised and unsupervised learning algorithms respectively are used for the disambiguation process. The Naive Bayes classifier achieved $92\%$ disambiguation accuracy. while the clustering performance of the EM algorithm is $67\%$ on the average. The retrieval effectiveness of the semantic vector space model incorporating the Naive Bayes classifier showed $39.6\%$ precision achieving about $7.4\%$ improvement. However, the retrieval effectiveness of the EM algorithm-based semantic retrieval is $3\%$ lower than the baseline retrieval without disambiguation. It is worth noting that the performances of disambiguation and retrieval depend on the distribution patterns of homonyms to be disambiguated as well as the characteristics of queries.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.1
• /
• pp.95-99
• /
• 2010

Social media refers to the content, which are created by users, such as blogs, social networks, and wikis. Recently, question-answering (QA) communities, in which users share information by questions and answers, are regarded as a kind of social media. Thus, QA communities have become a huge source of information for the past decade. However, it is hard for users to search the exact question-answer that is exactly matched with their needs as the number of question-answers increases in QA communities. This paper proposes an approach for classifying a question into three categories (information, opinion, and suggestion) according to the purpose of the question for more accurate information retrieval. Specifically, our approach is based on modified Na$\ddot{i}$ve Bayes classifier which uses structural characteristics of QA documents to improve the classification accuracy. Through our experiments, we achieved about 71.2% in classification accuracy.

• Journal of Digital Contents Society
• /
• v.14 no.4
• /
• pp.581-587
• /
• 2013

In this paper, we propose a method of sentiment classification which uses Levenshtein distance. We generate BOW(Bag-Of-Word) applying Levenshtein daistance in sentiment features and used it as the training set. Then the machine learning algorithms we used were SVMs(Support Vector Machines) and NB(Naive Bayes). As the data set, we gather 2,385 reviews of movies from an online movie community (Daum movie service). From the collected reviews, we pick sentiment words up manually and sorted 778 words. In the experiment, we perform the machine learning using previously generated BOW which was applied Levenshtein distance in sentiment words and then we evaluate the performance of classifier by a method, 10-fold-cross validation. As the result of evaluation, we got 85.46% using Multinomial Naive Bayes as the accuracy when the Levenshtein distance was 3. According to the result of the experiment, we proved that it is less affected to performance of the classification in spelling errors in documents.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2018.05a
• /
• pp.112-112
• /
• 2018

최근 이상기후로 인한 국부적인 혹은 광역적인 가뭄이 빈번하게 발생하고 있는 추세이며 발생횟수 뿐 아니라 가뭄 심도 및 지속기간이 과거보다 크게 증가하여 그에 따른 피해가 커질 것으로 예측되고 있다. 특히, 2014~2015년도의 유례없는 가뭄으로 인해 저수지 용수공급이 제한되면서 많은 농가들이 피해를 입었다. 본 연구의 목적은 전국 농업용 저수지를 대상으로 기상청 3개월 예보자료를 활용 할 수 있는 농업용 저수지 저수율 다중선형 회귀 모형을 개발하여 저수율 전망정보를 생산하는 것이다. 본 연구에서는 전국에 적용 가능한 저수율 다중선형 회귀 모형개발을 위해 5개의 기상요소(강수량, 최고기온, 최저기온, 평균기온, 평균풍속)와 관측 저수지 저수율을 활용했다. 기상자료는 2002년부터 2017년까지의 기상청 63개 지상관측소로부터 기상관측자료를 수집하였다. 본 연구에서는 저수율 전망 단계를 세 단계로 나누었다. 첫 번째 단계로 농어촌공사에서 전국 511개 용수구역을 대상으로 군집분석 및 의사결정나무 분석을 통해 제시한 65개 대표저수지를 대상으로 기상자료 및 관측 저수율 자료를 이용하여 다중선형 회귀분석을 실시하였다. 수집한 기상요소와 저수율을 독립변수로 하여 월별 회귀식을 산정한 결과 결정계수($R^2$)는 0.51~0.95로 나타났다. 두 번째 단계로 대표저수지의 회귀분석 결과를 전국의 저수지로 확대하기 위해 나이브 베이즈 분류법을 적용하여 전국 3098개의 저수지를 65의 군집으로 분류하고 각각의 군집에 해당되는 월별 회귀식을 산정하였다. 마지막으로 전국 저수지로 산정된 회귀식과 농업 가뭄 예측을 위해 기상청의 GS5(Global Seasonal Forecasting System 5) 3개월 예보자료를 수집하여 회귀식에 적용해 2017년 전국 저수지의 3개월 저수율 전망정보를 생산하였다. 본 연구의 전국 저수지 군집결과 기반의 저수율 전망기술은 2017년도 관측 저수율과 비교한 결과 유의한 상관성을 나타냈으며 이 결과는 추후 농업용 저수지의 물 공급 및 농업가뭄 전망 자료로서 이용이 가능할 것으로 판단된다.

• The Journal of the Korea Contents Association
• /
• v.16 no.10
• /
• pp.393-402
• /
• 2016

To compensate for limited the satisfaction survey currently conducted by Korea Metrological Administration (KMA), a sentiment analysis via a social networking service (SNS) can be utilized. From 2011 to 2014, with the sentiment analysis, Twitter who had commented 'KMA' had collected, then, using $Na{\ddot{i}}ve$ Bayes classification, we were classified into three sentiments: positive, negative, and neutral sentiments. An additional dictionary was made with morphemes appeared only in the positive, negative, and neutral sentiments of basic $Na{\ddot{i}}ve$ Bayes classification, thus the accuracy of sentiment analysis was improved. As a result, when sentiments were classified with a basic $Na{\ddot{i}}ve$ Bayes classification, the training data were reproduced about 75% accuracy rate. Whereas, when classifying with the additional dictionary, it showed 97% accuracy rate. When using the additional dictionary, sentiments of verification data was classified with about 75% accuracy rate. Lower classification accuracy rate would be improved by not only a qualified dictionary that has increased amount of training data, including diverse keywords related to weather, but continuous update of the dictionary. Meanwhile, contrary to the sentiment analysis based on dictionary definition of individual vocabulary, if sentiments are classified into meaning of sentence, increased rate of negative sentiment and change in satisfaction could be explained. Therefore, the sentiment analysis via SNS would be considered as useful tool for complementing surveys in the future.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.11a
• /
• pp.721-723
• /
• 2022

최근 낸드 플래시 메모리 기반의 Solid State Drive(SSD)가 기존 Hard Disk Drive(HDD)를 대신하여 개인용과 산업용으로도 널리 쓰이고 있다. 핫 데이터 구분 기법은 이러한 SSD 의 성능과 수명에 중요한 역할을 하는 Garbage Collection(GC)과 Wear Leveling(WL) 기술의 기반이 된다. 본 논문에서는 핫 데이터를 예측하기 위한 나이브 베이즈 분류 기반의 새로운 핫 데이터 구분 기법을 제안한다. 제안 기법은 워크로드 액세스 패턴의 학습 단계인 초기 단계와 실제 운영 단계를 통해 다시 액세스 될 확률이 높은 데이터를 그렇지 않은 데이터와 효과적으로 구분한다. 다양한 실제 trace 기반 실험을 통해 본 제안 기법이 기존 대표적인 기법보다 평균 19.3% 높은 성능을 확인했다.

• The KIPS Transactions:PartD
• /
• v.13D no.3 s.106
• /
• pp.369-376
• /
• 2006

This paper presents a new method of significantly improving conventional Bayesian statistical text classifier by incorporating accelerated EM(Expectation Maximization) algorithm. EM algorithm experiences a slow convergence and performance degrade in its iterative process, especially when real online-textual documents do not follow EM's assumptions. In this study, we propose a new accelerated EM algorithm with uncertainty-based selective sampling, which is simple yet has a fast convergence speed and allow to estimate a more accurate classification model on Naive Bayesian text classifier. Experiments using the popular Reuters-21578 document collection showed that the proposed algorithm effectively improves classification accuracy.