• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제10권11호
• /
• pp.5583-5604
• /
• 2016

In this paper, we present a novel image operator to extract textual information in natural scene images. First, a powerful refiner called the Stroke Color Extension, which extends the widely used Stroke Width Transform by incorporating color information of strokes, is proposed to achieve significantly enhanced performance on intra-character connection and non-character removal. Second, a character classifier is trained by using gradient features. The classifier not only eliminates non-character components but also remains a large number of characters. Third, an effective extractor called the Character Color Transform combines color information of characters and geometry features. It is used to extract potential characters which are not correctly extracted in previous steps. Fourth, a Convolutional Neural Network model is used to verify text candidates, improving the performance of text detection. The proposed technique is tested on two public datasets, i.e., ICDAR2011 dataset and ICDAR2013 dataset. The experimental results show that our approach achieves state-of-the-art performance.

• 정보관리학회지
• /
• 제39권1호
• /
• pp.91-117
• /
• 2022

국내를 비롯하여 전 세계적으로 우울증 환자 수가 매년 증가하는 추세이다. 그러나 대다수의 정신질환 환자들은 자신이 질병을 앓고 있다는 사실을 인식하지 못해서 적절한 치료가 이루어지지 않고 있다. 우울 증상이 방치되면 자살과 불안, 기타 심리적인 문제로 발전될 수 있기에 우울증의 조기 발견과 치료는 정신건강 증진에 있어 매우 중요하다. 이러한 문제점을 개선하기 위해 본 연구에서는 한국어 소셜 미디어 텍스트를 활용한 딥러닝 기반의 우울 경향 모델을 제시하였다. 네이버 지식인, 네이버 블로그, 하이닥, 트위터에서 데이터수집을 한 뒤 DSM-5 주요 우울 장애 진단 기준을 활용하여 우울 증상 개수에 따라 클래스를 구분하여 주석을 달았다. 이후 구축한 말뭉치의 클래스 별 특성을 살펴보고자 TF-IDF 분석과 동시 출현 단어 분석을 실시하였다. 또한, 다양한 텍스트 특징을 활용하여 우울 경향 분류 모델을 생성하기 위해 단어 임베딩과 사전 기반 감성 분석, LDA 토픽 모델링을 수행하였다. 이를 통해 문헌 별로 임베딩된 텍스트와 감성 점수, 토픽 번호를 산출하여 텍스트 특징으로 사용하였다. 그 결과 임베딩된 텍스트에 문서의 감성 점수와 토픽을 모두 결합하여 KorBERT 알고리즘을 기반으로 우울 경향을 분류하였을 때 가장 높은 정확률인 83.28%를 달성하는 것을 확인하였다. 본 연구는 다양한 텍스트 특징을 활용하여 보다 성능이 개선된 한국어 우울 경향 분류 모델을 구축함에 따라, 한국 온라인 커뮤니티 이용자 중 잠재적인 우울증 환자를 조기에 발견해 빠른 치료 및 예방이 가능하도록 하여 한국 사회의 정신건강 증진에 도움을 줄 수 있는 기반을 마련했다는 점에서 의의를 지닌다.

• International Journal of Contents
• /
• 제2권1호
• /
• pp.22-28
• /
• 2006

In this paper, we propose new semantic contents modeling using individual features, associative relations and visual features for efficiently supporting browsing and retrieval of video semantic contents. And we implement and design a browsing and retrieval system based on the semantic contents modeling. The browsing system supports annotation based information, keyframe based visual information, associative relations, and text based semantic information using a tree based browsing technique. The retrieval system supports text based retrieval, visual feature and associative relations according to the retrieval types of semantic contents.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제6권5호
• /
• pp.1373-1387
• /
• 2012

Contextual advertising is an important revenue source for major service providers on the Web. Ads classification is one of main tasks in contextual advertising, and it is used to retrieve semantically relevant ads with respect to the content of web pages. However, it is difficult for traditional text classification methods to achieve satisfactory performance in ads classification due to scarce term features in ads. In this paper, we propose a novel ads classification method that handles the lack of term features for classifying ads with short text. The proposed method utilizes a vocabulary expansion technique using semantic associations among terms learned from large-scale search query logs. The evaluation results show that our methodology achieves 4.0% ~ 9.7% improvements in terms of the hierarchical f-measure over the baseline classifiers without vocabulary expansion.

• 전자공학회논문지B
• /
• 제33B권10호
• /
• pp.120-129
• /
• 1996

This paper presents an efficient block segmentation and classification using the edge information of the document image. We extract four prominent features form the edge gradient and orientaton, all of which, and thereby the block clssifications, are insensitive to the background noise and the brightness variation of of the image. Using these four features, we can efficiently classify a document image into the seven categrories of blocks of small-size letters, large-size letters, tables, equations, flow-charts, graphs, and photographs, the first five of which are text blocks which are character-recognizable, and the last two are non-character blocks. By introducing the clumn interval and text line intervals of the document in the determination of th erun length of CRLA (constrained run length algorithm), we can obtain an efficient block segmentation with reduced memory size. The simulation results show that the proposed algorithm can rigidly segment and classify the blocks of the documents into the above mentioned seven categories and classification performance is high enough for all the categories except for the graphs with too much variations.

• Journal of Information Processing Systems
• /
• 제8권2호
• /
• pp.301-314
• /
• 2012

In this paper, two new approaches to align English-Arabic sentences in bilingual parallel corpora based on the Multi-Class Support Vector Machine (MSVM) and the Hidden Markov Model (HMM) classifiers are presented. A feature vector is extracted from the text pair that is under consideration. This vector contains text features such as length, punctuation score, and cognate score values. A set of manually prepared training data was assigned to train the Multi-Class Support Vector Machine and Hidden Markov Model. Another set of data was used for testing. The results of the MSVM and HMM outperform the results of the length based approach. Moreover these new approaches are valid for any language pairs and are quite flexible since the feature vector may contain less, more, or different features, such as a lexical matching feature and Hanzi characters in Japanese-Chinese texts, than the ones used in the current research.

• 대한산업공학회지
• /
• 제41권4호
• /
• pp.381-388
• /
• 2015

The rapid growth of the World Wide Web and online information services has generated and made accessible a huge number of text documents. To analyze texts, selecting important keywords is an essential step. In this paper, we propose a feature selection method that combines a term frequency-inverse document frequency technique and symmetrical conditional probability. The proposed method can identify features with N-gram, the sequential multiword. The effectiveness of the proposed method is demonstrated through a real text data from the machine learning repository, University of California, Irvine.

• 전자공학회논문지B
• /
• 제31B권10호
• /
• pp.82-91
• /
• 1994

In this thesis, we perform the experiment of speaker recognition by identifying vowels in the pronounciation of each speaker. In detail, we extract the vowels from the pronounciation of each speaker first. From it, we check the frequency energgy of 29 channels. After changing these into fuzzy values, we employ the fuzzy inference to recognize the speaker by text-dependent and text-independent methods. For this experiment, an algorithm of extracting vowels is developed, and newly introduced parameter is the frequency energy of the 29 channels computed from the extracted vowels. It shows the features of each speakers better than existing parameters. The advanced point of this paramter is to use the reference pattern only without the help of any codebook. As a rewult, test-dependent method showed about 95.5% rate of recognition, and text-independent method showed about 94.2% rate of recognition.

• 한국IT서비스학회지
• /
• 제11권sup호
• /
• pp.139-152
• /
• 2012

In the medical web forum, people share medical experience and information as patients and patents' families. Some people search medical information written in non-expert language and some people offer words of comport to who are suffering from diseases. Medical web forums play a role of the informative support and the emotional support. We propose the automatic classification model of articles in the medical web forum into the information support and emotional support. We extract text features of articles in web forum using text mining techniques from the perspective of linguistics and then perform supervised learning to classify texts into the information support and the emotional support types. We adopt the Support Vector Machine (SVM), Naive-Bayesian, decision tree for automatic classification. We apply the proposed model to the HealthBoards forum, which is also one of the largest and most dynamic medical web forum.

• 대한전기학회논문지:시스템및제어부문D
• /
• 제51권2호
• /
• pp.69-77
• /
• 2002

Speaker verification can be classified in two categories, text-dependent speaker verification and text-independent speaker verification. In this paper, we discuss text-dependent speaker verification. Text-dependent speaker verification system determines whether the sound characteristics of the speaker are equal to those of the specific person or not. In this paper we obtain the speaker data using a sound card in various noisy conditions, apply a new Hidden LMS (Least Mean Square) adaptive algorithm to it, and extract LPC (Linear Predictive Coding)-cepstrum coefficients as feature vectors. Finally, we use a competitive learning neural network for speaker verification. The proposed hidden LMS adaptive filter using a neural network reduces noise and enhances features in various noisy conditions. We construct a separate neural network for each speaker, which makes it unnecessary to train the whole network for a new added speaker and makes the system expansion easy. We experimentally prove that the proposed method improves the speaker verification performance.