• Journal of the Korea Society of Computer and Information
• /
• v.25 no.5
• /
• pp.187-197
• /
• 2020

Recently, various methods of text embedding using deep learning algorithms have been proposed. Especially, the way of using pre-trained language model which uses tremendous amount of text data in training is mainly applied for embedding new text data. However, traditional pre-trained language model has some limitations that it is hard to understand unique context of new text data when the text has too many tokens. In this paper, we propose self-supervised learning-based fine tuning method for pre-trained language model to infer vectors of long-text. Also, we applied our method to news articles and classified them into categories and compared classification accuracy with traditional models. As a result, it was confirmed that the vector generated by the proposed model more accurately expresses the inherent characteristics of the document than the vectors generated by the traditional models.

• Journal of Science and Technology Studies
• /
• v.12 no.1
• /
• pp.145-183
• /
• 2012

Science reporting is easier to lead to failure than other news reporting because it needs higher professionalism. According to Actor-Network Theory(ANT), not only research results(artifacts) of scientists but also science articles are hybrid networks. Namely, they are connected by human actors(scientist, reporter, etc.) and nonhuman actors(press releases etc.). When the process of science reporting is examined on the view of ANT, it is the process that scientists' results translate the media via press releases as intermediaries and expand their network to the public. This study aims at making an ANT analysis of how research results of Electronics and Telecommunications Research Institute(ETRI) scientist Hyun-Tak Kim were reported by lots of media, focusing on the rhetoric of ETRI's press release. It can reveal the reason for the science reporting's failure and hint at the better science journalism.

• Journal of Internet Computing and Services
• /
• v.22 no.5
• /
• pp.79-86
• /
• 2021

As the usefulness of big data analysis has been drawing attention, many studies in the business research area begin to use big data to predict firm performance. Previous studies mainly rely on data outside of the firm through news articles and social media platforms. The voices within the firm in the form of employee satisfaction or evaluation of the strength and weakness of the firm can potentially affect firm value. However, there is insufficient evidence that online employee reviews are valid to predict firm value because the data is relatively difficult to obtain. To fill this gap, from 2014 to 2019, we employed 97,216 reviews collected by JobPlanet, an online firm review website in Korea, and developed a machine learning-based predictive model. Among the proposed models, the LSTM-based model showed the highest accuracy at 73.2%, and the MAE showed the lowest error at 0.359. We expect that this study can be a useful case in the field of firm value prediction on domestic companies.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.6
• /
• pp.167-172
• /
• 2021

The objective of this study is to present the academic implications and developmental direction of the police assigned for special guard system through big data analysis on the objective and macroscopic viewpoint of the media. As research method, this study conducted the analysis on 'police assigned for special guard' and the analysis of related words that would visualize the keywords highly related to keyword trend and news. Also, after dividing the period into the 1990s, 2000s, and 2010s, the number of relevant articles in each period was drawn for understanding the flow. In the results of this study, the perception of media report of police assigned for special guard was about the recruitment of police assigned for special guard, and relevant events/accidents, which showed the coexistence of positive interest in the recruitment of police assigned for special guard and negative image of events/accidents related to police assigned for special guard. As a result, however, the necessity and demand for police assigned for special guard are increasing. Thus, the police assigned for special guard should be engaged in work after carefully thinking of its role in charge of ethical responsibility and safety as an axis for maintaining the national safety and social order.

• Journal of the Korea Society of Computer and Information
• /
• v.25 no.12
• /
• pp.35-44
• /
• 2020

In this study, we analyze how stock trend determination affects trend prediction accuracy. In stock markets, successful investment requires accurate stock price trend prediction. Therefore, a volume of research has been conducted to improve the trend prediction accuracy. For example, information extracted from SNS (social networking service) and news articles by text mining algorithms is used to enhance the prediction accuracy. Moreover, various machine learning algorithms have been utilized. However, stock trend determination has not been properly analyzed, and conventionally used methods have been employed repeatedly. For this reason, we formulate the trend determination as a moving average-based procedure and analyze its impact on stock trend prediction accuracy. The analysis reveals that trend determination makes prediction accuracy vary as much as 47% and that prediction accuracy is proportional to and inversely proportional to reference window size and target window size, respectively.

• Smart Media Journal
• /
• v.11 no.8
• /
• pp.73-83
• /
• 2022

In order to support new construction engineering companies with weak information power to successfully advance into the overseas construction market, this study tried to analyze what are the keywords of interest in the overseas construction market and how they differ from Korea. For this purpose, we recently collected 2,473 news article titles and major articles targeting smart cities that are of high interest in Korea and Vietnam. Through network configuration and topic modeling, we examined the connection relationship between the word of interest and the word of interest. In addition, the influence of the word of interest in the network was measured using PageRank centrality. Through this analysis, it was found that there is a high interest in smart city-related construction, cities, and digital in both countries, and the difference in terms of interest between Korea and Vietnam was inferred. Finally, the limitations of this study and additional research directions to complement them are presented.

• Smart Media Journal
• /
• v.12 no.5
• /
• pp.17-27
• /
• 2023

Recently, automatic text summarization, which automatically summarizes only meaningful information for users, is being studied steadily. Especially, research on text summarization using Transformer, an artificial neural network model, has been mainly conducted. Among various studies, the GSG method, which trains a model through sentence-by-sentence masking, has received the most attention. However, the traditional GSG has limitations in selecting a sentence to be masked based on the degree of overlap of tokens, not the meaning of a sentence. Therefore, in this study, in order to improve the quality of text summarization, we propose SbGSG (Semantic-based GSG) methodology that selects sentences to be masked by GSG considering the meaning of sentences. As a result of conducting an experiment using 370,000 news articles and 21,600 summaries and reports, it was confirmed that the proposed methodology, SbGSG, showed superior performance compared to the traditional GSG in terms of ROUGE and BERT Score.

• The Research Journal of the Costume Culture
• /
• v.32 no.3
• /
• pp.381-397
• /
• 2024

This study investigates the negative perceptions (veganphobia) held by consumers toward vegan diets and fashion and aims to foster a genuine acceptance of ethical veganism in consumption. The textual data web-crawled Korean online posts, including news articles, blogs, forums, and tweets, containing keywords such as "contradiction," "dilemma," "conflict," "issues," "vegan food" and "vegan fashion" from 2013 to 2021. Data analysis was conducted through text mining, network analysis, and clustering analysis using Python and NodeXL programs. The analysis revealed distinct negative perceptions regarding vegan food. Key issues included the perception of hypocrisy among vegetarians, associations with specific political leanings, conflicts between environmental and animal rights, and contradictions between views on companion animals and livestock. Regarding the vegan fashion industry, the eco-friendliness of material selection and design processes were seen as the pivotal factors shaping negative attitudes. Furthermore, the study identified a shared negative perception regarding vegan food and vegan fashion. This negativity was characterized by confusion and conflicts between animal and environmental rights, biased perceptions linked to specific political affiliations, perceived self-righteousness among vegetarians, and general discomfort toward them. These factors collectively contributed to a broader negative perception of vegan consumption. In conclusion, this study is significant in understanding the complex perceptions and attitudes that con- sumers hold toward vegan food and fashion. The insights gained from this research can aid in the design of more effective campaign strategies aimed at promoting vegan consumerism, ultimately contributing to a more widespread acceptance of ethical veganism in society.

• Korean journal of communication and information
• /
• v.46
• /
• pp.449-484
• /
• 2009

It has been said that Gwangjudream, nevertheless a free press, plays a role as a local press that it should be, in a situation that other local papers do not. This study aims to reveal the forces that influence the article production of Gwangjudream, and to examine the interrelations between them, through using the methods of participant observations and depth interviews. In this course, it is eventually purpose of providing more deep understandings on the present circumstances and problems of the local papers and having a chance to concern the concrete ways to enhance them. This study results in revealing the five forces that primarily influence the article production of Gwangjudream: 1) as a historical force, keeping the spirit of the first publication that look forward to playing a role as a local press that it sound be, 2) as an individual force, the habitus of its members that is critical of mainstream society and culture, 3) as an organizational force, non-hierarchical culture and the independence of the editorial rights, 4) as a habitual force, the deny of beat system, 5) as an economical force, the power of sponsors, financial poorness, and the competition for attracting subscribers. While the historical force and the individual force play a role as fundamental circumstances and the organizational force and the habitual force as practical circumstances for producing articles, they encourage to emerge the characteristics of the articles that are related to citizens' everyday life and reflect locality, and criticize and keep an eye on government and other public offices. However, the economical force provides the circumstances that weaken the characteristics of Gwangjudream. The results of this study question the perspective to overly regard it as coming from their economical weakness that the local newspaper do not play a role as a local press that it should be.

• Journal of Internet Computing and Services
• /
• v.22 no.2
• /
• pp.109-121
• /
• 2021

As the global COVID-19 pandemic is prolonged, the Corona Blue phenomenon, combined with COVID-19 and blue, is intensifying. The purpose of this study is to analyze the current trend of Corona Blue in consideration of the possibility of increasing mental illness and the need for countermeasures, especially after COVID-19. This study tried to find out the relationship between stress and leisure activities before and after COVID-19 by using Corona Blue news article analysis through the topic modeling method, and questionnaire find out the help of stress and leisure activities. This study was compared and analyzed using two research methods. First, a total of 363 news articles were analyzed through topic modeling based on newspaper articles from January 2020, when COVID- 19 was upgraded to the "border" stage, until September, where the social distancing stage was strengthened to stage 2.5 in Korea. As a result of the study, a total of 28 topics were extracted, and similar topics were grouped into 7 groups: mental-demic, generational spread, causes of depression acceleration, increased fatigue, attitude to coping with long-term wars, changes in consumption, and efforts to overcome depression. Second, the SPSS statistical program was used to analyze the level of stress change according to leisure activities before/after COVID-19 and the main help according to leisure activities. As a result of the study, it was confirmed that the average difference in stress reduction according to participation in leisure activities before COVID-19 was larger than after COVID-19. Also, leisure activities were found to be effective in stress relief even after COVID-19. In addition, if the main help from leisure activities before COVID-19 was the meaning of relaxation and recharging through physical and social activities. After COVID-19, psychological roles such as mood swings through nature, outdoor activities, or intellectual activities were found to play a large part. As such, in this study, it was confirmed that understanding the current status of Corona Blue and coping with leisure in extreme stress situations has a positive effect. It is expected that this research can serve as a basis for preparing realistic and desirable leisure policies and countermeasures to overcome Corona Blue.