• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.11
• /
• pp.4426-4442
• /
• 2020

Since scene segmentation is becoming a hot topic in the field of autonomous driving and medical image analysis, researchers are actively trying new methods to improve segmentation accuracy. At present, the main issues in image semantic segmentation are intra-class inconsistency and inter-class indistinction. From our analysis, the lack of global information as well as macroscopic discrimination on the object are the two main reasons. In this paper, we propose a Densely connected residual Attention network (DA-Res2Net) which consists of a dense residual network and channel attention guidance module to deal with these problems and improve the accuracy of image segmentation. Specifically, in order to make the extracted features equipped with stronger multi-scale characteristics, a densely connected residual network is proposed as a feature extractor. Furthermore, to improve the representativeness of each channel feature, we design a Channel-Attention-Guide module to make the model focusing on the high-level semantic features and low-level location features simultaneously. Experimental results show that the method achieves significant performance on various datasets. Compared to other state-of-the-art methods, the proposed method reaches the mean IOU accuracy of 83.2% on PASCAL VOC 2012 and 79.7% on Cityscapes dataset, respectively.

• Journal of Korean Elementary Science Education
• /
• v.37 no.1
• /
• pp.39-53
• /
• 2018

The purpose of this study was to investigate the conceptual structure of subjects related to matter having pre-service elementary school teachers by applying semantic network analysis (SNA). The analyzed concepts in the subjects of matter were 6 words such as 'atom', 'molecule', 'ion', 'electron', 'matter' and 'particle'. The results of SNA of the concepts are as follows : 1. In the semantic network of 'atom', words having a high betweenness centrality were linked with the words based on both the scientific context and the everyday context. 2. The network of 'molecule' was analyzed to be more organized than the network of the 'atom'. 3. In the network of 'ion', the group of words of the scientific context was distinguished from the group of words of the everyday context. 4. The network of 'electron' was analyzed to be more oriented on electricity and magnetism in the field of physics. 5. In the network of 'matter', the words related to compounds were linked with knowledge of history of science. 6. The network of 'particle' was not structured with words based on particulate nature of matter.

• Journal of the Korean Society of Clothing and Textiles
• /
• v.42 no.3
• /
• pp.428-437
• /
• 2018

This study derives evaluation terms by analyzing the semantic relationship between design elements and sentiment terms in regards to fashion design. As for research methods, a total of 38,225 texts from Daum and Naver Blogs from November 2015 to October 2016 were collected to analyze the parts, frequency, centrality and semantic networks of the terms. As a result, design elements were derived in the form of a noun while fashion image and user's emotional responses were derived in the form of adjectives. The study selected 15 noun terms and 52 adjective terms as evaluation terms for men's striped shirts. The results of semantic network analysis also showed that the main contents of the users of men's striped shirts were derived as characteristics of expression, daily wear, formation, and function. In addition, design elements such as pattern, color, coordination, style, and fit were classified with evaluation results such as wide, bright, trendy, casual, and slim.

• Journal of Gifted/Talented Education
• /
• v.27 no.1
• /
• pp.17-35
• /
• 2017

The purpose of this study is to analyze the learning objectives characteristics of educational programs of centers for the university affiliated science-gifted education using semantic network analysis, we examined the applicability of semantic network analysis in analyzing learning objectives by comparing the results of analysis with Bloom's revised taxonomy. For this purpose, 702 learning objectives presented in 169 science subjects were selected as subjects to be analyzed. After classifying and coding the learning objectives according to Bloom's revised taxonomy, we conducted a semantic network analysis to investigate the relationship between learning objectives. The results of the analysis are as follows. First, we looked at the number of learning objectives used for each subject, and about 3 elementary school levels and about 6 middle school levels were used. Second, the knowledge dimension such as 'factual and conceptual knowledge' and cognitive process dimension such as 'remember', 'understand', and 'create' was high regardless of the research method and school level. Third, the results of analysis based on the weighting through the semantic network analysis method, the elementary school level emphasize activities th be applied to the actual experimental process through learning about scientific facts, while the middle school level emphasize the understanding of scientific facts and concepts themselves. As a result, it can be seen that the semantic network analysis can analyze characteristics of various learning objectives rather than the conventional simple statistical analysis.

• Journal of Korean Clinical Nursing Research
• /
• v.25 no.3
• /
• pp.209-220
• /
• 2019

Purpose: A significant amount of public opinion about nurse bullying is expressed on the internet. The purpose of this study was to analyze the linkage structures among words extracted from comments on internet articles related to nurse workplace bullying using semantic network analysis. Methods: From February 2018 to April 2019, comments made on news articles posted to the Daum and Naver web portal containing keywords such as "nurse", "Taeum", and "bullying" were collected using a web crawler written in Python. A morphological analysis performed with Open Korean Text in KoNLPy generated 54 major nodes. The frequencies, eigenvector centralities, and betweenness centralities of the 54 nodes were calculated and semantic networks were visualized using the UCINET and NetDraw programs. Convergence of iterated correlations (CONCOR) analysis was performed to identify structural equivalence. Results: This paper presents results about March 2018 and January 2019 because these months had highest number of articles. Of the 54 major nodes, "nurse", "hospital", "patient", and "physician" were the most frequent and had the highest eigenvector and betweenness centralities. The CONCOR analysis identified work environment, nurse, gender, and military clusters. Conclusion: This study structurally explored public opinion about nurse bullying through semantic network analysis. It is suggested that various studies on nursing phenomena will be conducted using social network analysis.

• Journal of International Academy of Physical Therapy Research
• /
• v.10 no.4
• /
• pp.1926-1933
• /
• 2019

Background: Physical therapy has been widely studied in various fields, however, the academic trends and characteristics has not been systematically analyzed. Semantic network analysis is used as an approach for this study. Objective: To explore academic trends and knowledge system in the physiotherapy research in the Journal of International Academy Physical Therapy (J of IAPTR) Study design : Literature review Method: Semantic network analysis was conducted using the titles of 272 articles published in the Journal of IAPTR from 2010 to 2019. Results: Frequency analysis revealed following most frequently used key words; Stroke (27 times), Balance (21 times), Elder (13 times), Forward head posture (FHP, 11 times), Muscle activity (9 times). The relationship between the presented keywords is divided into six subgroups (FHP and pain, walk and quality, elder and balance, stroke and apoptosis, muscle strength and function) according to their correlation and frequency to be used together. Conclusion: The study is considered to be of help to researchers who want to identify research trends in physiotherapy.

• Journal of Korea Multimedia Society
• /
• v.21 no.12
• /
• pp.1425-1430
• /
• 2018

In this paper, we propose a CNN based deep learning algorithm for semantic segmentation of images. In order to improve the accuracy of semantic segmentation, we combined pixel level object classification and image level object classification. The image level object classification is used to accurately detect the characteristics of an image, and the pixel level object classification is used to indicate which object area is included in each pixel. The proposed network structure consists of three parts in total. A part for extracting the features of the image, a part for outputting the final result in the resolution size of the original image, and a part for performing the image level object classification. Loss functions exist for image level and pixel level classification, respectively. Image-level object classification uses KL-Divergence and pixel level object classification uses cross-entropy. In addition, it combines the layer of the resolution of the network extracting the features and the network of the resolution to secure the position information of the lost feature and the information of the boundary of the object due to the pooling operation.

• CELLMED
• /
• v.13 no.2
• /
• pp.3.1-3.21
• /
• 2023

In this study, the drinking culture of the Choseon dynasty is examined with the text frequency analysis technique on the entire 『Choseonwangjosilok (朝鮮王朝實錄)』. This study examined a total of 1,968 volumes and 948 books about 27 kings of Choseon , which spans a total of 518 years, through web crawling on the National Institute of Korean History website. Python 3.8 was used to extract sentences related to alcohol, Rhino 1.4.5 was used for morphological analysis to extract nouns, and Gephi 0.9.2 was used for semantic network analysis. According to 『Choseonwangjosilok (朝鮮王朝實錄)』 about alcohol culture, the results of the analysis are as follow: Alcoholic beverages were more often used in court or in ritual ceremonies rather than those based on specific ingredients or manufacturing methods commonly used by the general public. regarding the ruling class through semantic network analysis l in the 『Choseonwangjosilok (朝鮮王朝實錄)』, the Choseon dynasty was found to be highly associated with political issues related to maintaining the power relations within the Korean royal court system. At times, alcohol was used to maintain personal relationships, while at other times it was seen as an essential item in state ceremonies. It was also used as a highly political means to maintain and strengthen national power.

• The Research Journal of the Costume Culture
• /
• v.30 no.5
• /
• pp.766-778
• /
• 2022

Image-oriented information is becoming increasingly important on social networking services (SNS); the background of this trend is the popularity of selfies. Currently, camera applications using augmented reality (AR) and artificial intelligence (AI) technologies are gaining traction. An AR camera app is a smartphone application that converts selfies into various interesting forms using filters. In this study, we investigated the change of keywords according to the time flow of selfies in Goolgle News articles through semantic network analysis. Additionally, we examined the effects of using an AR camera app on appearance satisfaction and self-esteem when taking a selfie. Semantic network analysis revealed that in 2013, postings of specific people were the most prominent selfie-related keywords. In 2019, keywords appeared regarding the launch of a new smartphone with a rear-facing camera for selfies; in 2020, keywords related to communication through selfies appeared. As a result of examining the effect of the degree of use of the AR camera app on appearance satisfaction, it was found that the higher the degree of use, the higher the user's interest in appearance. As a result of examining the effect of the degree of use of the AR camera app on self-esteem, it was found that the higher the degree of use, the higher the user's negative self-esteem.

• Journal of Intelligence and Information Systems
• /
• v.10 no.2
• /
• pp.1-13
• /
• 2004

This paper presents a method for the multilingual product retrieval agent through XML and the semantic networks in e-commerce. Retrieval for products is an important process, since it represents interfaces of the customer contact to the e-commerce. Keyword-based retrieval is efficient as long as the product information is structured and organized. But when the product information is expressed across many online shopping malls, especially when it is expressed in different languages with cultural backgrounds, buyers' product retrieval needs language translation with ambiguities resolved in a specific context. This paper presents a RDF modeling case that resolves semantic problems in the representation of product information and across the boundaries of language domains. With adoption of UNSPSC code system, this paper designs and implements an architecture for the multilingual product retrieval agents. The architecture is based on the central repository model of product catalog management with distributed updating processes. It also includes the perspectives of buyers and suppliers. And the consistency and version management of product information are controlled by UNSPSC code system. The multilingual product names are resolved by semantic networks, thesaurus and ontology dictionary for product names.