• Journal of Digital Convergence
• /
• v.19 no.6
• /
• pp.251-258
• /
• 2021

In this paper, we proposed an encoder-decoder model utilizing residual learning to improve the accuracy of the U-Net-based semantic segmentation method. U-Net is a deep learning-based semantic segmentation method and is mainly used in applications such as autonomous vehicles and medical image analysis. The conventional U-Net occurs loss in feature compression process due to the shallow structure of the encoder. The loss of features causes a lack of context information necessary for classifying objects and has a problem of reducing segmentation accuracy. To improve this, The proposed method efficiently extracted context information through an encoder using residual learning, which is effective in preventing feature loss and gradient vanishing problems in the conventional U-Net. Furthermore, we reduced down-sampling operations in the encoder to reduce the loss of spatial information included in the feature maps. The proposed method showed an improved segmentation result of about 12% compared to the conventional U-Net in the Cityscapes dataset experiment.

• The KIPS Transactions:PartA
• /
• v.12A no.6 s.96
• /
• pp.539-546
• /
• 2005

Recently, as the number of Internet users are growing explosively, e-learning has been applied spread, as well as remote evaluation of intellectual capacity However, only the multiple choice and/or the objective tests have been applied to the e-learning, because of difficulty of natural language processing. For the intelligent marking of short-essay typed answer papers with rapidness and fairness, this work utilize heterogenous linguistic knowledges. Firstly, we construct the semantic kernel from un tagged corpus. Then the answer papers of students and instructors are transformed into the vector form. Finally, we evaluate the similarity between the papers by using the semantic kernel and decide whether the answer paper is correct or not, based on the similarity values. For the construction of the semantic kernel, we used latent semantic analysis based on the vector space model. Further we try to reduce the problem of information shortage, by integrating Korean Word Net. For the construction of the semantic kernel we collected 38,727 newspaper articles and extracted 75,175 indexed terms. In the experiment, about 0.894 correlation coefficient value, between the marking results from this system and the human instructors, was acquired.

• Culinary science and hospitality research
• /
• v.24 no.3
• /
• pp.167-176
• /
• 2018

In this study, Big Data was used to collect the information related to 'Cooking Academy' keywords. After collecting all the data, we calculated the frequency through the text mining and selected the main words for future data analysis. Data collection was conducted from Google Web and News during the period from January 1, 2013 to December 31, 2017. The selected 64 words were analyzed by using UCINET 6.0 program, and the analysis results were visualized with NetDraw in order to present the relationship of main words. As a result, it was found that the most important goal for the students from cooking school is to work as a cook, likewise to have practical classes. In addition, we obtained the result that SNS marketing system that the social sites, such as Facebook, Twitter, and Instagram are actively utilized as a marketing strategy of the institute. Therefore, the results can be helpful in searching for the method of utilizing big data and can bring brand-new ideas for the follow-up studies. In practical terms, it will be remarkable material about the future marketing directions and various programs that are improved by the detailed curriculums through semantic network of cooking school by using big data.

• International Journal of Contents
• /
• v.15 no.1
• /
• pp.10-24
• /
• 2019

This study applied network agenda-setting theory to analyze the impact of the agenda-setting function of the media on certain issues by focusing on the agenda at the center of controversy, 'Creative Economy'. To this end, the study extracted the data referred to creative economy in the media and SNS from 1 January 2008 to 31 December 2014, and analyzed the data using the network analysis program UCINET and the Korean language analysis program Textom. The results of the present study show that, during the period under former President Lee (2008-2011), the media's creative economy agenda-setting function did not exert a significant impact on the agenda-setting within SNS. However, from 2012 when the government of former President Park Geun-hye had started, the agenda-setting function of the media starts to show increasingly strong influence on the agenda cognition in SNS. The central words and sub-words configuration forming the center of the semantic network moved in the direction of a high correlation, in addition to the gradually increasing correlation based on QAP correlation analysis. In 2014, the semantic networks of the media and SNS bore a close resemblance to each other, while the shape of networks and sub-words structure also had a high level of similarity.

• Journal of the Regional Association of Architectural Institute of Korea
• /
• v.21 no.5
• /
• pp.109-116
• /
• 2019

The Seoul city recommended that old apartments be preserved, and as part of that, it decided to preserve some of the buildings for Jamsil Jugong, which was built in 1977, and Gaepo Jugong, which was constructed in 1981. The purpose of this study was to compare and review newspaper articles with two perspectives positive and negative about how the social perception of 'apartment trace remaining' was being constructed. By looking at the meaning of keywords delivered by newspaper articles and the interaction structure between keywords through the analysis of semantic networks, we analyzed how the media is pursuing an issue on the topic of preservation of architectural cultural heritage. The analysis results confirmed that there was a clear difference between positive and negative newspaper. Positive articles dealt with utilization from the point of view of keywords linked to preservation, and negative articles showed that keywords related to the property and backlash of residents linked to the policy of the Seoul Metropolitan Government were linked, leading to high negative public opinion.

• International Journal of Computer Science & Network Security
• /
• v.23 no.9
• /
• pp.120-128
• /
• 2023

In this work, we have presented ATSA, a hierarchical attention deep learning model for Arabic sentiment analysis. ATSA was proposed by addressing several challenges and limitations that arise when applying the classical models to perform opinion mining in Arabic. Arabic-specific challenges including the morphological complexity and language sparsity were addressed by modeling semantic composition at the Arabic morphological analysis after performing tokenization. ATSA proposed to perform phrase-chunks sentiment embedding to provide a broader set of features that cover syntactic, semantic, and sentiment information. We used phrase structure parser to generate syntactic parse trees that are used as a reference for ATSA. This allowed modeling semantic and sentiment composition following the natural order in which words and phrase-chunks are combined in a sentence. The proposed model was evaluated on three Arabic corpora that correspond to different genres (newswire, online comments, and tweets) and different writing styles (MSA and dialectal Arabic). Experiments showed that each of the proposed contributions in ATSA was able to achieve significant improvement. The combination of all contributions, which makes up for the complete ATSA model, was able to improve the classification accuracy by 3% and 2% on Tweets and Hotel reviews datasets, respectively, compared to the existing models.

• Journal of Information Technology Applications and Management
• /
• v.30 no.5
• /
• pp.1-19
• /
• 2023

This study examines Korean users' attitudes and emotions toward Melon and Spotify, which lead the music streaming market. We used Text Mining, Semantic Network Analysis, TF-IDF, Centrality, CONCOR, and Word2Vec analysis. As a result of the study, MelOn was used in a user's daily life. Based on Melon's advantages of providing various contents, the advantage is judged to have considerable competitiveness beyond the limits of the streaming app. However, the MelOn users had negative emotions such as anger, repulsion, and pressure. On the contrary, in the case of Spotify, users were highly interested in the music content. In particular, interest in foreign music was high, and users were also interested in stock investment. In addition, positive emotions such as interest and pleasure were higher than MelOn users, which could be interpreted as providing attractive services to Korean users. While previous studies have mainly focused on technical or personal factors, this study focuses on consumer reactions (online reviews) according to corporate strategies, and this point is the differentiation from others.

• Journal of the Korean Society for information Management
• /
• v.26 no.2
• /
• pp.173-194
• /
• 2009

Recently, semantic search techniques which are based on information space as consisting of nonambiguous, non-redundant, formal pieces of ontological knowledge have been developed so that users do exploit large knowledge bases. The purpose of the study is to design more user-friendly and smarter retrieval interface based on ontological analysis, which can provide more precise information by reducing semantic ambiguity or more rich linked information based on well-defined relationships. Therefore, this study, first of all, focuses on ontological analysis on researcher information as selecting descriptive elements, defining classes and properties of descriptive elements, and identifying relationships between the properties and their restriction between relationships. Next, the study designs the prototypical retrieval interface based on ontology-based representation, which supports to semantic searching and browsing regarding researchers as a full-fledged domain. On the proposed retrieval interface, users can search various facts for researcher information such as research outputs or the personal information, or carrier history and browse the social connection of the researchers such as researcher group that is lecturing or researching on the same subject or involving in the same intellectual communication.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.23 no.1
• /
• pp.1-6
• /
• 2019

This study visualized users reaction about movies based on keywords with high frequency. For this work, we collected data of movie reviews on . A total of six movies were selected, and we conducted the work of data gathering and preprocessing. Semantic network analysis was used to understand the relationship among keywords. Also, NetDraw, packaged with UCINET, was used for data visualization. In this study, we identified the differences in characteristics of review contents regarding each movie. The implication of this study is that we visualized movie reviews made by sentence as keywords and explored whether it is possible to construct the interface to check users' reaction at a glance. We suggest that further studies use more diverse movie reviews, and the number of reviews for each movie is used in similar quantities for research.

• Journal of the Korean BIBLIA Society for library and Information Science
• /
• v.30 no.1
• /
• pp.181-207
• /
• 2019

This study analyzed social issues related to presidential records in press releases using semantic network analysis method. For this purpose, we 1) selected five major news medias in Korea - Chosun Ilbo, JoongAng Ilbo, Dong-A Ilbo, Hankyoreh, and Kyunghyang Newspaper; 2) collected relevant articles including the subject word "Presidential Records", and 3) analyzed issue trends based on timeline using semantic network. According to medias, the issue related to the presidential records were analyzed by comparing the specific keywords in terms of persons, entities, actions. At the results, It is possible to identify the reporting patterns and components of the presidential records related issues. And the difference of media characteristics according to news media tendency was derived.