• The Journal of Information Systems
• /
• v.26 no.4
• /
• pp.285-301
• /
• 2017

Purpose In this paper we verified the reliabilities of healthcare-related information provided by various users on the site of Naver Jisikin, a Korean typical search platform. Based on Q&A contents we validated answers' reliabilities to the asked questions about a lung cancer with the help of professors at a medical school. Design/methodology/approach The content analysis includes that the types of questions are classified into symptom/diagnosis, therapy, prognosis, after-management and so on. The answers contains advice, advertisement, oriental medicine, and religion as well as the above 5 question categories. The validation results of medical evidence about each answer show that only 49% among all answers have medical grounds. Findings We classified the medical grounded answers into three levels; high, medium and low. Among all answers we need to find out the answers including advertisement because the answers can be harmful to patients. We found the method to select the answers containing advertisement contents with the help of text mining research. The selection model presents high performance as 84% classification accuracy.

• Language and Information
• /
• v.14 no.2
• /
• pp.159-181
• /
• 2010

This paper presents a hybrid approach to the sentiment analysis of online texts. The sentiment of a text refers to the feelings that the author of a text has towards a certain topic. Many existing approaches employ either a pattern-based approach or a machine learning based approach. The former shows relatively high precision in classifying the sentiments, but suffers from the data sparseness problem, i.e. the lack of patterns. The latter approach shows relatively lower precision, but 100% recall. The approach presented in the current work adopts the merits of both approaches. It combines the pattern-based approach with the machine learning based approach, so that the relatively high precision and high recall can be maintained. Our experiment shows that the hybrid approach improves the F-measure score for more than 50% in comparison with the pattern-based approach and for around 1% comparing with the machine learning based approach. The numerical improvement from the machine learning based approach might not seem to be quite encouraging, but the fact that in the current approach not only the sentiment or the polarity information of sentences but also the additional information such as target of sentiments can be classified makes the current approach promising.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.11 no.5
• /
• pp.193-201
• /
• 2011

This paper presents the holistic notion and model of an open social u-learning community, anchored with open content, providing an interactive online study group experience akin to sitting with study buddies on a world-wide campus quad. The interactive social u-learning community design helps conceptualize and maximize advantages of ubiquitous environment in learning. The model is enabled by state-of-the-art web technologies; real-time collaboration technologies for a highly interactive experience; intelligent recommender systems to help learners connect with relevant content and other learners; and mining and analytics to assess learner outcomes. Hence, u-learning design is highly scalable yet interactive and engaging.

• Journal of Preventive Medicine and Public Health
• /
• v.53 no.3
• /
• pp.171-174
• /
• 2020

The coronavirus disease 2019 (COVID-19) pandemic has not only caused significant challenges for health systems all over the globe but also fueled the surge of numerous rumors, hoaxes, and misinformation, regarding the etiology, outcomes, prevention, and cure of the disease. Such spread of misinformation is masking healthy behaviors and promoting erroneous practices that increase the spread of the virus and ultimately result in poor physical and mental health outcomes among individuals. Myriad incidents of mishaps caused by these rumors have been reported globally. To address this issue, the frontline healthcare providers should be equipped with the most recent research findings and accurate information. The mass media, healthcare organization, community-based organizations, and other important stakeholders should build strategic partnerships and launch common platforms for disseminating authentic public health messages. Also, advanced technologies like natural language processing or data mining approaches should be applied in the detection and removal of online content with no scientific basis from all social media platforms. Furthermore, these practices should be controlled with regulatory and law enforcement measures alongside ensuring telemedicine-based services providing accurate information on COVID-19.

• Korean Journal of Child Studies
• /
• v.37 no.1
• /
• pp.7-16
• /
• 2016

Objective: It aimed at proposing the Korean Journal of Child Studies' strategy to be promoted to international journal based on the style and format of scholarly journals and journal metrics. Methods: The review of the journal in not only print version, but also an online version was done from the perspective of style and format. The total citation and impact factor were manually calculated from Web of Science Core Collection. Results: More professional level manuscript editing is required for maintaining the consistency of the style and format. The verso page and back matters should be improved to international level. Journal homepage should be reconstructed by adopting digital standards for the journal, including journal article tag suite, CrossMark, FundRef, ORCID, and text and data mining. To become an international journal, transformation into English journal and deposition to PubMed Central is mandatory. Conclusion: Since the editor's and society members' performance is top-notch, it will be possible to promote the journal up to international level soon. Society should guarantee the term of editor for enough time and support her with full cost and complete consent.

• Journal of the Korea Society of Computer and Information
• /
• v.13 no.1
• /
• pp.37-44
• /
• 2008

Data stream has the tendency to change in Patterns over time. Also known as concept drift, such problem can reduce the predictive performance of a classification model CVFDT and IOLIN tried to solve the problem of a concept drift through incremental classification model updates. The local changes in patterns. however was revealed to be unable to resolve the problems of local concept drift that occurs by influencing on total classification results. In this paper, we propose adapted IOLIN system that improves system's predictive performance by detecting the local concept drift. The experimental result shows that adaptive IOLIN, the Proposed method, is about 2.8% in accuracy better than IOLIN and about 11.2% in accuracy better than CVFDT.

• Asia Pacific Journal of Business Review
• /
• v.1 no.1
• /
• pp.1-22
• /
• 2016

The prediction of prices of agricultural products in the agriculture IT sector plays a significant role in the economic life of consumers and anyone engaged in agricultural business, and as these prices fluctuate more often than do other prices, the prediction of these prices holds a great deal of research promise. For this reason, academic literature has provided studies on the factors influencing the prices of agricultural products and the price index. However, as these factors vary, they are difficult to predict, resulting in the challenge of acquiring quantitative data. China is one example of a country without a reliable prediction system for prices of agricultural products. Fortunately, disclosed heterogeneous data can be found on the Internet, which allows for the effective collection of factors related to the prediction of these product prices through the use of text mining. The data provided online is valuable in that they reflect the opinions of the general public in real-time. Accordingly, this study aims to use heterogeneous data from the Internet and suggest a model predicting the prices of agricultural products before functional analyses. Toward this end, data analyses were conducted on the Chinese agricultural products market, one of the largest markets in the world.

• Journal of Korea Game Society
• /
• v.15 no.3
• /
• pp.177-188
• /
• 2015

In-game cheating by the use of unauthorized software programs has always been a big problem that they can damage in First Person Shooting games, although companies operate a variety of client security solutions in order to prevent games from the cheating attempts. This paper proposes a method for detecting cheaters in FPS games by using game log analysis in a server-side. To accomplish this, we did a comparative analysis of characteristics between cheaters and general users focused on commonly loaded logs in the game. We proposed a cheating detection model by using artificial neural network algorithm. In addition, we did the performance evaluation of the proposed model by using the real dataset used in business.

• The KIPS Transactions:PartD
• /
• v.15D no.1
• /
• pp.139-146
• /
• 2008

It is very important to identify useful web user group and induce their behavior pattern in eCRM domain. Inducing user group with a similar inclination, a reliability of user group decreases because there is an uncertainty in online user data. In this paper, we have applied the 2-tier clustering, which uses the outcome of interaction with data from other tiers. Also we propose a method which induces user behavior pattern from a cluster and compare C4.5 with our method.

• Journal of Information Processing Systems
• /
• v.16 no.1
• /
• pp.201-209
• /
• 2020

The graph data structure is popular because it can intuitively represent real-world knowledge. Graph databases have attracted attention in academia and industry because they can be used to maintain graph data and allow users to mine knowledge. Mining reachability relationships between two nodes in a graph, termed reachability query processing, is an important functionality of graph databases. Online traversals, such as the breadth-first and depth-first search, are inefficient in processing reachability queries when dealing with large-scale graphs. Labeling schemes have been proposed to overcome these disadvantages. The state-of-the-art is the 2-hop labeling scheme: each node has in and out labels containing reachable node IDs as integers. Unfortunately, existing 2-hop labeling schemes generate huge 2-hop label sizes because they only consider local features, such as degrees. In this paper, we propose a more efficient 2-hop label size reduction approach. We consider the topological sort index, which is a global feature. A linear combination is suggested for utilizing both local and global features. We conduct experiments over real-world and synthetic directed acyclic graph datasets and show that the proposed approach generates smaller labels than existing approaches.