• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.2
• /
• pp.308-313
• /
• 2021

In this study, tweets about the 2019 Koreas-US DMZ Summit were collected along with a time sequence and analyzed by a sequential topic modeling method, Dynamic Topic Modeling(DTM). In microblogging services such as Twitter, unstructured data that mixes news and an opinion about a single event occurs at the same time on a large scale, and information and reactions are produced in the same message format. Therefore, to grasp a topic trend, the contextual meaning can be found only by performing pattern analysis reflecting the characteristics of sequential data. As a result of calculating the DTM after obtaining the topic coherence score and evaluating the Latent Dirichlet Allocation(LDA), 30 topics related to news reports and opinions were derived, and the probability of occurrence of each topic and keywords were dynamically evolving. In conclusion, the study found that DTM is a suitable model for analyzing the trend of integrated topics in a specific event over time.

• The Journal of the Korea Contents Association
• /
• v.21 no.7
• /
• pp.1-12
• /
• 2021

This study explored the mechanism by which stars and fans build a global fandom community based on social connectivity of social media based on BTS' official Twitter and in-depth interviews with Chinese, American, and Korean fans. Stars and fans use the connectivity of social media to build a pseudo-private relationship that crosses public and private affairs. Even though they won the award, BTS "congratulates" the fan club, ARMY, and organizes a dialogue method so that each fan becomes the subject. From a fan's point of view, it comes to be seen as a personal message to them, which is not just applied to the text composition method. On days when there is no public message such as official activities or anniversaries, a device to communicate without missing a day has been prepared by raising private messages or previous memories. It was found that the sense of constant connection strengthens fans' sense of community toward each other.

• Journal of Technology Innovation
• /
• v.26 no.1
• /
• pp.31-57
• /
• 2018

The general public's interest level towards safer life is increasing due to not only ever-changing faces of disasters and increased frequency of climate-change related disasters but also enhanced standard of living. Demand for disaster-safety industry is also increasing. Several policies for disaster-safety industry have been introduced. The policies, however, did not fully reflect the level of people's interest. This study is to investigate possible ways to reflect general public's interests towards disaster-safety industry using social media analysis, so that disaster-safety industry can be properly promoted. To examine the level of general public's interest, social media data during the last three years were compiled and analyzed. It was found that the interest level was highest towards, firstly, information on just-happened real disasters, secondly, necessary knowledge in real life which could be applied immediately if disasters strike. It was also confirmed that social media was useful in analyzing people's interest level quickly, because social data have been found to be sharply increased during the 2016 Gyeongju Earthquake in Korea. This study suggests applicable plans for disaster-related industry promotion based on social media data using general public's interest level.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.11 no.1
• /
• pp.76-81
• /
• 2018

The Internet of Things (IoT), which is emerging as a future economic growth engine, has been actively introduced in areas close to our daily lives. However, there are still IoT security threats that need to be resolved. In particular, with the spread of smart homes and smart cities, an explosive amount of closed-circuit televisions (CCTVs) have been installed. The Internet protocol (IP) information and even port numbers assigned to CCTVs are open to the public via search engines of web portals or on social media platforms, such as Facebook and Twitter; even with simple tools these pieces of information can be easily hacked. For this reason, a big-data analytics system is needed, capable of supporting quick responses against data, that can potentially contain risk factors to security or illegal websites that may cause social problems, by assisting in analyzing data collected by search engines and social media platforms, frequently utilized by Internet users, as well as data on illegal websites.

• KIPS Transactions on Software and Data Engineering
• /
• v.6 no.6
• /
• pp.309-314
• /
• 2017

In general, there are many difficulties in collecting, storing, and analyzing SNS (social network service) data, since those data have big data characteristics, which occurs very fast with the mixture form of structured and unstructured data. In this paper, we propose a new data visualization framework that works on Apache Storm, and it can be useful for real-time and dynamic analysis of SNS data. Apache Storm is a representative big data software platform that processes and analyzes real-time streaming data in the distributed environment. Using Storm, in this paper we collect and aggregate the real-time Twitter data and dynamically visualize the aggregated results through the tag cloud. In addition to Storm-based collection and aggregation functionalities, we also design and implement a Web interface that a user gives his/her interesting keywords and confirms the visualization result of tag cloud related to the given keywords. We finally empirically show that this study makes users be able to intuitively figure out the change of the interested subject on SNS data and the visualized results be applied to many other services such as thematic trend analysis, product recommendation, and customer needs identification.

• Journal of Digital Convergence
• /
• v.13 no.4
• /
• pp.395-404
• /
• 2015

This study is exploring objective awareness of literary therapy by consideration of popular perception about literary therapy through analysis of big data. The purpose of this study is the deduction of meaning information through analysis in the viewpoint of big data at online social network service(SNS) about 'literary therapy'. Accordingly, the main way of research became content analysis of keyword linked to literary therapy by utilizing opinion mining method related to text mining. The study mainly grasped 'literary therapy' and analyzed 'bibliotherapy' comparatively. The period of study was from Oct. 10th to Nov. 10th, 2014(during 30 days), and SNS such as blog or twitter became the subject of search. Through the result of study analysis, the conclusion that the spread of literary therapeutic prospect, structural harmony of literary therapeutic field, and the solidity of perceptional axis about literary therapy are needed can be drawn. This study is worthwhile because it can investigate popular awareness about literary therapy and can suggest alternative for invigoration of literary therapy.

• The Journal of the Korea Contents Association
• /
• v.12 no.7
• /
• pp.463-472
• /
• 2012

For last decade, the amount of information has been increased rapidly because of the internet and computing technology development, mobile devices and sensors, and social networks like facebook or twitter. People who want to gain important knowledge from database have been frustrated with large database. Many studies for automatic knowledge extracting meaningful knowledge from large database have been fulfilled. In that sense, automatic knowledge extracting with computing technology has been highly significant in information technology field, but still has many challenges to go further. In order to improve the effectives and efficiency of knowledge extracting system, test collection is strongly necessary. In this research, we introduce a test collection for automatic knwoledge extracting. We name the test collection KEEC/KREC(KISTI Entity Extraction Collection/KISTI Relation Extraction Collection) and present the process and guideline for building as well as the features of. The main feature is to tag by experts to guarantee the quality of collection. The experts read documents and tag entities and relation between entities with a tool for tagging. KEEC/KREC is being used for a research to evaluate system performance and will continue to contribute to next researches.

• Journal of the Korea Society for Simulation
• /
• v.29 no.4
• /
• pp.9-19
• /
• 2020

The number of customers on Social Network Services(SNS) is rapidly increasing with the spread of smartphones. As of 2018, about 2.7 billion people of the world population (about 7.1 billion people) and more than 31.2 million people of the total population of South Korea (about 50.1 million) use SNS. There are several studies have been conducted on increasing SNS market. Most of them, however, were not quantitative but qualitative studies. This study is conducted on domestic SNS market to identify the competitive relationship among SNS platforms with great proportion in South Korea, such as Facebook, Instagram and Twitter. The objective is to suggest some hypotheses of the competitive relations, test them, and finally verify the trend of domestic SNS market. Competitive Lotka-Volterra (LV) model is used to find out the competitive relationships and Moving Window is also used to show the changes of them over time. In order to test the hypotheses on the relationships, some experiments are performed with Moving Window technique. Thus, the relations among the platforms and the changes of them over time are identified.

• Journal of the Society of Disaster Information
• /
• v.18 no.4
• /
• pp.687-698
• /
• 2022

Purpose: This study was conducted to understand the public awareness of landslide and check dams in 2015-2020 using the big data platform 'Hyean' and to confirm the utilization of this platform in disaster prevention areas. Method: The total amount, number of detection by period by media, and affirmative and negative trends of a search for 'landslide' and 'check dam' in 2015-2020 were analyzed using a keyword search of 'Hyean.' Result: There is significant lack of public awareness of check dam compared to landslide, and the trend is more noticeable in the conspicuous gap of data amount between the news and SNS media. The number and the timing of the search for 'landslide' coincided with the actual occurrence of landslide, while the detection of 'check dam' was less related to it. Relatively affirmative preception for the check dam is inferred, but it was difficult to confirm accurate statistical affirmative and negative trends in the disaster prevention field using 'Hyean.' Conclusion: Unlike the experts who expect positive public awareness of check dam, the statistic results show that the public awareness of the check dam as an effective countermeasure against landslide was extremely low. Active promotion of erosion control projects should be carried out first, and a balanced sample survey should accompany online and periodic field surveys. Since there is a limit to grasping the effective perception in the field of disaster prevention area using 'Hyean', it should be very cautious to establish local/governmental policies using it.

• Journal of Internet Computing and Services
• /
• v.24 no.4
• /
• pp.127-135
• /
• 2023

This research paper explores the application of FinBERT, a variational BERT-based model pre-trained on financial domain, for sentiment analysis in the financial domain while focusing on the process of identifying suitable training data and hyperparameters. Our goal is to offer a comprehensive guide on effectively utilizing the FinBERT model for accurate sentiment analysis by employing various datasets and fine-tuning hyperparameters. We outline the architecture and workflow of the proposed approach for fine-tuning the FinBERT model in this study, emphasizing the performance of various datasets and hyperparameters for sentiment analysis tasks. Additionally, we verify the reliability of GPT-3 as a suitable annotator by using it for sentiment labeling tasks. Our results show that the fine-tuned FinBERT model excels across a range of datasets and that the optimal combination is a learning rate of 5e-5 and a batch size of 64, which perform consistently well across all datasets. Furthermore, based on the significant performance improvement of the FinBERT model with our Twitter data in general domain compared to our news data in general domain, we also express uncertainty about the model being further pre-trained only on financial news data. We simplify the complex process of determining the optimal approach to the FinBERT model and provide guidelines for selecting additional training datasets and hyperparameters within the fine-tuning process of financial sentiment analysis models.