• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.11 no.8
• /
• pp.3021-3027
• /
• 2010

Example-based chatBot generates a response to user's utterance by searching the most similar utterance in a collection of dialogue examples. Though finding an appropriate example is very important as it is closely related to a response quality, few studies have reported regarding what features should be considered and how to use the features for similar utterance searching. In this paper, we propose a machine learning framework which uses various linguistic features. Experimental results show that simultaneously using both semantic features and lexical features significantly improves the performance, compared to conventional approaches, in terms of 1) the utilization of example database, 2) precision of example matching, and 3) the quality of responses.

• Korean Institute of Interior Design Journal
• /
• v.22 no.1
• /
• pp.183-193
• /
• 2013

The study aimed to find the fundamental differences of aesthetics in Korea, China, and Japan by analyzing expressive features and structural meanings of the storage furniture from $17^{th}$ to $19^{th}$ century. The study was performed in four steps; analysis of expressive features, isotopic analysis, semantic structure analysis, and comprehensive interpretation. The results showed that three countries had linear shapes with curvilinear elements, closed forms with open spaces, natural material hues with change of tone or color, and symmetrical forms with asymmetrical patterns and structures in common. Korea comparatively accented on the natural material colors and wood grains. China stressed on the big and wide faces using three-dimensional carving. Japan accented on the linear elements with strong color contrast and decorative metal fixtures. These features were caused by the traditional thoughts and according aesthetic principles. Korea and China were affected by the Confucianism focused on establishing the order of rank. Meanwhile, Japan was more influenced by the Buddhism emphasized on the individuality and communication. Therefore, the differences of the expressive features in furniture among the three countries were inevitable consequences of the different ideologies.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.16 no.10
• /
• pp.3390-3405
• /
• 2022

Long Short-Term Memory (LSTM) combined with attention mechanism is extensively used to generate semantic sentences of images in image captioning models. However, features of salient regions and spatial information are not utilized sufficiently in most related works. Meanwhile, the LSTM also suffers from the problem of underutilized information in a single time step. In the paper, two innovative approaches are proposed to solve these problems. First, the Synergy-Gated Attention (SGA) method is proposed, which can process the spatial features and the salient region features of given images simultaneously. SGA establishes a gated mechanism through the global features to guide the interaction of information between these two features. Then, the Recurrent Fusion LSTM (RF-LSTM) mechanism is proposed, which can predict the next hidden vectors in one time step and improve linguistic coherence by fusing future information. Experimental results on the benchmark dataset of MSCOCO show that compared with the state-of-the-art methods, the proposed method can improve the performance of image captioning model, and achieve competitive performance on multiple evaluation indicators.

• Journal of Digital Contents Society
• /
• v.12 no.2
• /
• pp.241-252
• /
• 2011

During the last decades, other parts of operating systems, storage devices, and media are changed steadily, whereas filesystem is changed little. As data is grown bigger, the number of files to be managed also increases in geometrically. Researches about new filesystem schemes are being done widely to support these files efficiently. In web document search area, there are many researches about finding meaningful documents using semantic search. Many researches tried to apply these schemes, which is been proven in web document search previously, to filesystems. But they've focused only on higher layer of filesystem, that is not related seriously to storage media. Therefore they're not well tuned to physical characteristics of new flash memory based SSD which has different features against traditional HDD. We enhance log structured filesystem, that is already well known to work better in SSD, by putting semantic search scheme to and with multi logging point.

• Journal of KIISE:Databases
• /
• v.34 no.6
• /
• pp.554-563
• /
• 2007

The e-Marketplace is growing rapidly and providing a more complex relationship between providers and consumers. In recent years, e-Marketplace integration or cooperation issues have become an important issue in e-Business. The e-Catalog is a key factor in e-Business, which means an e-Catalog System needs to contain more large data and requires a more efficient retrieval system. This paper focuses on designing an efficient retrieval system for very large e-Catalogs of large e-Marketplaces. For this reason, a new similarity measure for e-Catalog retrieval based on semantic relationships was proposed. Our achievement is this: first, a new e-Catalog data model based on semantic relationships was designed. Second, the model was extended by considering lexical features (Especially, focus on Korean). Third, the factors affecting similarity with the model was defined. Fourth, from the factors, we finally defined a new similarity measure, realized the system and verified it through experimentation.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.30 no.4 s.152
• /
• pp.531-541
• /
• 2006

The purpose of this study is to examine the semantic structure about the image shown on the product, brand and advertisement and figuring out its features through the correlation among brand images. For the study, nine brands(Guess, Bangbang, ONG, NIX, TBJ, Levi's, OPT, FRJ, Jambangee) as subjects for investigation has been selected and divided into the image of brand(9 brands), product(108 products, 12 pieces for each product) and advertisement(9 points) by the measure of 26 adjective pairs. The survey has been collected on the subject of 540 men and women who live in around Busan city areas and has been taken the statistics. The results on investigating the semantic structures of the product images about jean brands, there are five main factors, such as, individuality, attractiveness, activeness, modernity, hardness & softness. The results on examining the semantic structures of the brand images about jean brands, the factors are attractiveness, activeness, vitality, hardness & softness, fadness. The results on investigating the semantic structure of the advertisement images about jean brands, the factors are attractiveness, individuality, modernity, activeness. The results on the classification of brand images are presented as four groups, the first group is that brand and advertisement image are pretty similar but product image is differential according to brand and the second group, product and advertisement image are similar but brand image is differential. The third group is that product and brand image are similar but advertisement is differential and the fourth group, product, brand and advertisement are similar.

• Korean Journal of Remote Sensing
• /
• v.34 no.6_2
• /
• pp.1239-1249
• /
• 2018

Optical remote sensing sensors provide visually more familiar images than radar images. However, it is difficult to discriminate sea ice types in optical images using spectral information based machine learning algorithms. This study addresses two topics. First, we propose a semantic segmentation which is a part of the state-of-the-art deep learning algorithms to identify ice types by learning hierarchical and spatial features of sea ice. Second, we propose a new approach by combining of semi-supervised and active learning to obtain accurate and meaningful labels from unlabeled or unseen images to improve the performance of supervised classification for multiple images. Therefore, we successfully added new labels from unlabeled data to automatically update the semantic segmentation model. This should be noted that an operational system to generate ice type products from optical remote sensing data may be possible in the near future.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.5
• /
• pp.662-667
• /
• 2022

Semantic role labeling task(SRL) is to extract predicate and arguments such as agent, patient, place, time. In the previously SRL task studies, a pipeline method extracting linguistic features of sentence has been proposed, but in this method, errors of each extraction work in the pipeline affect semantic role labeling performance. Therefore, methods using End-to-End neural network model have recently been proposed. In this paper, we propose a neural network model using the Biaffine Average Attention model for SRL task. The proposed model consists of a structure that can focus on the entire sentence information regardless of the distance between the predicate in the sentence and the arguments, instead of LSTM model that uses the surrounding information for prediction of a specific token proposed in the previous studies. For evaluation, we used F1 scores to compare two models based BERT model that proposed in existing studies using F1 scores, and found that 76.21% performance was higher than comparison models.

• Journal of the Korea Convergence Society
• /
• v.12 no.11
• /
• pp.45-51
• /
• 2021

With the development of deep learning, semantic segmentation methods are being studied in various fields. There is a problem that segmenation accuracy drops in fields that require accuracy such as medical image analysis. In this paper, we improved PSPNet, which is a deep learning based segmentation method to minimized the loss of features during semantic segmentation. Conventional deep learning based segmentation methods result in lower resolution and loss of object features during feature extraction and compression. Due to these losses, the edge and the internal information of the object are lost, and there is a problem that the accuracy at the time of object segmentation is lowered. To solve these problems, we improved PSPNet, which is a semantic segmentation model. The multi-scale attention proposed to the conventional PSPNet was added to prevent feature loss of objects. The feature purification process was performed by applying the attention method to the conventional PPM module. By suppressing unnecessary feature information, eadg and texture information was improved. The proposed method trained on the Cityscapes dataset and use the segmentation index MIoU for quantitative evaluation. As a result of the experiment, the segmentation accuracy was improved by about 1.5% compared to the conventional PSPNet.

• The KIPS Transactions:PartB
• /
• v.13B no.2 s.105
• /
• pp.121-126
• /
• 2006

In this paper, we propose an approach to increase the power of clustering Web images by using high-level semantic features from text information relevant to Web images as well as low-level visual features of image itself. These high-level text features can be obtained from image URLs and file names, page titles, hyperlinks, and surrounding text. As a clustering engine, self-organizing map (SOM) proposed by Kohonen is used. In the SOM-based clustering using high-level text features and low-level visual features, the 200 images from 10 categories are divided in some suitable clusters effectively. For the evaluation of clustering powers, we propose simple but novel measures indicating the degrees of scattering images from the same category, and degrees of accumulation of the same category images. From the experiment results, we find that the high-level text features are more useful in SOM-based Web image clustering.