• The KIPS Transactions:PartB
• /
• v.13B no.4 s.107
• /
• pp.409-416
• /
• 2006

In this paper, we propose an algorithm for extracting text regions from spam-mail images using color layer. The CLTE(color layer-based text extraction) divides the input image into eight planes as color layers. It extracts connected components on the eight images, and then classifies them into text regions and non-text regions based on the component sizes. We also propose an algorithm for recovering damaged text strokes from the extracted text image. In the binary image, there are two types of damaged strokes: (1) middle strokes such as 'ㅣ' or 'ㅡ' are deleted, and (2) the first and/or last strokes such as 'ㅇ' or 'ㅁ' are filled with black pixels. An experiment with 200 spam-mail images shows that the proposed approach is more accurate than conventional methods by over 10%.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.18 no.6A
• /
• pp.129-137
• /
• 2008

As the use of the internet increases, the distribution of spam mail has also vastly increased. The email's main use was for the exchange of information, however, currently it is being more frequently used for advertisement and malware distribution. This is a serious problem because it consumes a large amount of the limited internet resources. Furthermore, an extensive amount of computer, network and human resources are consumed to prevent it. As a result much research is being done to prevent and filter spam. Currently, research is being done on readable sentences which do not use proper grammar. This type of spam can not be classified by previous vocabulary analysis or document classification methods. This paper proposes a method to filter spam by using the subject of the mail and N-GRAM for indexing and Bayesian, SVM algorithms for classification.

• Journal of KIISE:Software and Applications
• /
• v.32 no.5
• /
• pp.378-384
• /
• 2005

In this paper, we describe a novel method of spam-filtering to improve the performance of conventional spam-filtering systems. Conventional systems filter emails by investigating words distribution in email headers or bodies. Nowadays, spammers begin making email accounts in web-based email service sites and sending emails as if they are not spams. Investigating the email accounts of those spams, we notice that there is a large difference between the automatically generated accounts and ordinaries. Based on that difference, incoming emails are classified into spam/non-spam classes. To classify emails from only account strings, we used decision trees, which have been generally used for conventional pattern classification problems. We collected about 2.15 million account strings from email service sites, and our account checker resulted in the accuracy of $96.3\%$. The previous filter system with the checker yielded the improved filtering performance.

• Journal of KIISE:Software and Applications
• /
• v.29 no.4
• /
• pp.270-277
• /
• 2002

Text filtering is a task of deciding whether a document has relevance to a specified topic. As Internet and Web becomes wide-spread and the number of documents delivered by e-mail explosively grows the importance of text filtering increases as well. The aim of this paper is to improve the accuracy of text filtering systems by using machine learning techniques. We apply AdaBoost algorithms to the filtering task. An AdaBoost algorithm generates and combines a series of simple hypotheses. Each of the hypotheses decides the relevance of a document to a topic on the basis of whether or not the document includes a certain word. We begin with an existing AdaBoost algorithm which uses weak hypotheses with their output of 1 or -1. Then we extend the algorithm to use weak hypotheses with real-valued outputs which was proposed recently to improve error reduction rates and final filtering performance. Next, we attempt to achieve further improvement in the AdaBoost's performance by first setting weights randomly according to the continuous Poisson distribution, executing AdaBoost, repeating these steps several times, and then combining all the hypotheses learned. This has the effect of mitigating the ovefitting problem which may occur when learning from a small number of data. Experiments have been performed on the real document collections used in TREC-8, a well-established text retrieval contest. This dataset includes Financial Times articles from 1992 to 1994. The experimental results show that AdaBoost with real-valued hypotheses outperforms AdaBoost with binary-valued hypotheses, and that AdaBoost iterated with random weights further improves filtering accuracy. Comparison results of all the participants of the TREC-8 filtering task are also provided.

• The Journal of the Korea Contents Association
• /
• v.10 no.6
• /
• pp.43-48
• /
• 2010

Useful emails influence on consumers' purchase behavior and activate them to visit retail stores. Regular contact with consumers by e-mail has positive effects on brand loyalty. However, email marketing has a limitation. Spam now accounts for over half of all e-mail traffic. The increase of email users has resulted in the dramatic increase of spam emails during the past few years. In this paper, we proposed an ontology-based system offering personalized email services to overcome such limitation. Our method is not the ontology-driven spam filtering, but a personalized content service considering personal interests and relations among people by using FOAF and domain ontologies. Our system was successfully tested in email marketing domain.

• Journal of the Korea Society for Simulation
• /
• v.11 no.2
• /
• pp.1-16
• /
• 2002

The need for network security is being increasing due to the development of information communication and internet technology. In this paper, firewall models, operating system models and other network component models are constructed. Each model is defined by basic or compound model, referencing DEVS formalism. These models and the simulation environment are implemented with MODSIM III, a general purpose, modular, block-structured high-level programming language which provides direct support for object-oriented programming and discrete-event simulation. In this simulation environment with representative attacks, the following three attacks are generated, SYN flooding and Smurf attack as an attack type of denial of service, Mail bomb attack as an attack type of e-mail. The simulation is performed with the models that exploited various security policies against these attacks. The results of this study show that the modeling method of packet filtering system, proxy system, unix and windows NT operating system. In addition, the results of the simulation show that the analysis of security performance according to various security policies, and the analysis of correlation between availability and confidentiality according to security empowerment.

• Journal of KIISE:Computing Practices and Letters
• /
• v.9 no.5
• /
• pp.487-497
• /
• 2003

There have been many filtering systems covering mail or news. Documents filtered by them consist of general text or HTML. XML is emerging as a new standard for information exchange. So, filtering systems need new approaches in dealing with XML documents. Our system suggests a method to describe user profiles with XML's ability to represent schema and structure. An user profile is made from DTD information and it is supposed to point the specific part of a document conforming to the DTD. More, it is different from the existing systems in extracting part of a document. An user profile is reflected in XML query to get part of an XML document.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04c
• /
• pp.552-554
• /
• 2003

전자우편은 기존 우편 기능을 대체하는 대표적인 정보 전달 수단으로 자리 잡고 있다. 전자매일 사용자의 증가에 따라 망은 기업들은 전자 메일을 통해 광고를 하게 되었다. 이에 따라 전자매일 사용자들은 인터넷 상에 개인 전자메일 주소가 노출됨으로 많은 스팸메일을 수신하게 되는데, 이것은 전자메일 사용자에게 많은 부담이 되고있다. 본 논문은 전자우편 문서내의 단어들을 대상으로 통계적 방법의 SVM을 이용하여 스팸메일을 필터링 하였으며, 학습 단계에서 단어 자질공간의 축소를 위해 DF값 변화에 따른 학습을 통하여 분류의 성능을 비교하였다. SVM의 성능 평가를 위해 확률적 방법의 나이브 베이지안과 벡터 모텔을 이용한 분류기와 성능을 비교함으로써 SVM 방법이 우수한 성능을 보임을 검증하였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2003.04c
• /
• pp.549-551
• /
• 2003

본 논문은 점점 증가되고 있는 SPAM 메일 문제를 해결하기 위한 방법으로써, 특정 속성에 기반을 둔 학습 알고리즘의 co-training을 통한 전자메일 분류 기법을 제안한다. 전자메일 분류는 결국 문서 분류 기술과 다르지 않다. 이미 많은 연구에서 학습 알고리즘을 이용한 문서 분류 기법은 많이 제안되고 검증되었다. 본 논문에서는 이러한 학습 알고리즘들을 co-training을 통하여 해당 메일이 SPAM인지 아닌지 구분하며, 학습의 효율성을 높이기 위하여 전자메일의 특정한 속성들, 예를 들면, 핵심문구나 기타 특정한 문구 및 전자메일의 헤더 정보 등을 학습 기반으로 이용하였다.

• Journal of the Korea Computer Industry Society
• /
• v.5 no.4
• /
• pp.453-460
• /
• 2004

SPAM mails become a serious social problem all of the world and the products for SPAM prevention are coming to the market. This study classifies the existing SPAM prevention solutions according to the patterns to be set up and the judging SPAM methods, and analyses the merits and demerits of them. This study also draws problems of the existing SPAM Prevention solutions and suggests a new URL Prefetch method, a new filtering method which have been out of use. And it draws synergistic effects of SPAM prevention by this new method and suggests SPAM Prevention solution by HTML Pattern method