• Journal of Internet Computing and Services
• /
• v.24 no.5
• /
• pp.17-27
• /
• 2023

Today, as AI (Artificial Intelligence) technology develops and its practicality increases, it is widely used in various application fields in real life. At this time, the AI model is basically learned based on various statistical properties of the learning data and then distributed to the system, but unexpected changes in the data in a rapidly changing data situation cause a decrease in the model's performance. In particular, as it becomes important to find drift signals of deployed models in order to respond to new and unknown attacks that are constantly created in the security field, the need for lifecycle management of the entire model is gradually emerging. In general, it can be detected through performance changes in the model's accuracy and error rate (loss), but there are limitations in the usage environment in that an actual label for the model prediction result is required, and the detection of the point where the actual drift occurs is uncertain. there is. This is because the model's error rate is greatly influenced by various external environmental factors, model selection and parameter settings, and new input data, so it is necessary to precisely determine when actual drift in the data occurs based only on the corresponding value. There are limits to this. Therefore, this paper proposes a method to detect when actual drift occurs through an Anomaly analysis technique based on XAI (eXplainable Artificial Intelligence). As a result of testing a classification model that detects DGA (Domain Generation Algorithm), anomaly scores were extracted through the SHAP(Shapley Additive exPlanations) Value of the data after distribution, and as a result, it was confirmed that efficient drift point detection was possible.

• Journal of Internet Computing and Services
• /
• v.25 no.2
• /
• pp.79-92
• /
• 2024

ChatGPT, as a representative chatbot leveraging generative artificial intelligence technology, is used valuable not only in scientific and technological domains but also across diverse sectors such as society, economy, industry, and culture. This study conducts an explorative analysis of user sentiments and needs for ChatGPT by examining global social media discourse on Reddit. We collected 10,796 comments on Reddit from December 2022 to August 2023 and then employed keyword analysis, sentiment analysis, and need-mining-based topic modeling to derive insights. The analysis reveals several key findings. The most frequently mentioned term in ChatGPT-related comments is "time," indicative of users' emphasis on prompt responses, time efficiency, and enhanced productivity. Users express sentiments of trust and anticipation in ChatGPT, yet simultaneously articulate concerns and frustrations regarding its societal impact, including fears and anger. In addition, the topic modeling analysis identifies 14 topics, shedding light on potential user needs. Notably, users exhibit a keen interest in the educational applications of ChatGPT and its societal implications. Moreover, our investigation uncovers various user-driven topics related to ChatGPT, encompassing language models, jobs, information retrieval, healthcare applications, services, gaming, regulations, energy, and ethical concerns. In conclusion, this analysis provides insights into user perspectives, emphasizing the significance of understanding and addressing user needs. The identified application directions offer valuable guidance for enhancing existing products and services or planning the development of new service platforms.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.24 no.2
• /
• pp.167-175
• /
• 2024

Due to the Fourth Industrial Revolution, ICT(Information and Communication Technology) industry is becoming more important and sophisticated than ever. In B2B based ICT industry demand forecasting by analyzing the previous customer data is so important. RFM, one of customer relationship management models is a marketing technique that evaluates Recency, Frequency and Monetary value to predict customers behavior. RFM model has been studied focusing on the B2C based industry. On the other hand there is a lack of research on B2B based technology industry. Therefore this study applied it to B2B based high technology industry and considered T(technology collaboration) value, which are identified as important factors in the technology industry. To present an improved model for market performance in B2B technology industry, an empirical study was conducted on comparing the accuracy of the traditional RFM model and the improved RFM-T model. The objective of this study is to contribute to market performance by presenting an improved model in B2B based high technology industry.

• Journal of Internet Computing and Services
• /
• v.25 no.3
• /
• pp.43-54
• /
• 2024

This study employs text mining techniques to compare user responses across various Over-The-Top (OTT) service platforms. The primary objective of the research is to understand user satisfaction with OTT service platforms and contribute to the formulation of more effective review strategies. The key questions addressed in this study involve identifying prominent topics and keywords in user reviews of different OTT services and comprehending platform-specific user reactions. TF-IDF is utilized to extract significant words from positive and negative reviews, while BERTopic, an advanced topic modeling technique, is employed for a more nuanced and comprehensive analysis of intricate user reviews. The results from TF-IDF analysis reveal that positive app reviews exhibit a high frequency of content-related words, whereas negative reviews display a high frequency of words associated with potential issues during app usage. Through the utilization of BERTopic, we were able to extract keywords related to content diversity, app performance components, payment, and compatibility, by associating them with content attributes. This enabled us to verify that the distinguishing attributes of the platforms vary among themselves. The findings of this study offer significant insights into user behavior and preferences, which OTT service providers can leverage to improve user experience and satisfaction. We also anticipate that researchers exploring deep learning models will find our study results valuable for conducting analyses on user review text data.

• Asia pacific journal of information systems
• /
• v.18 no.3
• /
• pp.67-90
• /
• 2008

Prior to making choices among online products and services, consumers often search online product reviews written by other consumers. Online product reviews have great influences on consumer behavior because they are believed to be more reliable than information provided by sellers. However, ever-increasing lists of product reviews make it difficult for consumers to find the right information efficiently. A customized search mechanism is a method to provide personalized information which fits the user's requirements. This study examines effects of a customized search mechanism and perceived similarity between consumers and product reviewers on consumer behaviors. More specifically, we address the following research questions: (1) Can a customized search mechanism increase perceived similarity between product review authors and readers? (2) Are product reviews perceived as more credible when product reviews were written by the authors perceived similar to them? (3) Does credibility of product reviews have a positive impact on acceptance of product reviews? (4) Does acceptance of product reviews have an influence on purchase intention of the readers? To examine these research questions, a lab experiment with a between-subject factor (whether a customized search mechanism is provided or not) design was employed. In order to enhance mundane realism and increase generalizability of the findings, the experiment sites were built based on a real online store, cherrya.com (http://www.cherrya.com/). Sixty participants were drawn from a pool that consisted of undergraduate and graduate students in a large university. Participation was voluntary; all the participants received 5,000 won to encourage their motivation and involvement in the experiment tasks. In addition, 15 participants, who selected by a random draw, received 30,000 won to actually purchase the product that he or she decided to buy during the experiment. Of the 60 participants, 25 were male and 35 were female. In examining the homogeneity between the two groups, the results of t-tests revealed no significant difference in gender, age, academic years, online shopping experience, and Internet usage. To test our research model, we completed tests of the measurement models and the structural models using PLS Graph version 3.00. The analysis confirmed individual item reliability, internal consistency, and discriminant validity of measurements. The results show that participants feel more credible when product reviews were written by the authors perceived similar to them, credibility of product reviews have a positive impact on acceptance of product reviews, and acceptance of product reviews have an influence on purchase intention of the readers. However, a customized search mechanism did not increase perceived similarity between product review authors and readers. The results imply that there is an urgent need to develop a better customized search tool in order to increase perceived similarity between product review authors and readers.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.14 no.4
• /
• pp.63-76
• /
• 2011

This study purposes to develop a model to analyze soil loss in estimating prior disaster influence. The model of analyzing soil loss develops the soil loss analysis system on the basis of Internet by introducing cloud computing system, and also develops a standalone type in connection with HyGIS. The soil loss analysis system is developed to draw a distribution chart without requiring a S/W license as well as without preparing basic data such as DEM, soil map and land cover map. Besides, it can help users to draw a soil loss distribution chart by applying various factors like direct rain factors. The tools of Soil Loss Anaysis Model in connection with HyGiS are developed as add-on type of GMMap2009 in GEOMania, and also are developed to draw Soil Loss Hazard Map suggested by OECD. As a result of using both models, they are developed very conveniently to analyze soil loss. Hereafter, these models will be able to be improved continuously through researches to analyze sediment a watershed outlet and to calculate R value using data of many rain stations.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.1038-1044
• /
• 2009

Interests for User Interface directly influencing on usability are also growing as the competition between DCATV and IPTV come to the surface, according to getting activated Interactive TV market. Therefore, this study has the purpose to understand mental models of users for Interactive TV service & Interface. The methods for this study are as follows. First, as the step considering literature study, we compared differences between DCATV and IPTV and caught up the concepts of DCATV and IPTV among some types of Interactive TV. Second, we examinated and analysed Hello TV, C&M, Broad&TV, MegaTV, myLGtv as the step of analysis for examples for Interactive TV. Third, as the step of researches observing users we analysed usability problems caused by differences of mental models between DCATV and IPTV. As a result, we could find there are some differences for mental model of users in Interface & services between DCATV and IPTV Especially, it was found that users feel confused very much regarding perception of real time broadcasting and VOD. We regards this findings can be used as basic for concept makings of Interactive TV, when TV interface is developed in the future.

• The Korean Journal of Financial Management
• /
• v.25 no.1
• /
• pp.23-49
• /
• 2008

We test the hypothesis that the gradual diffusion of information across asset markets leads to cross-asset return predictability in Korea. And, the aim of this paper is related to forecast the stock market, business cycle index and industrial production by various indicators of economic activities in Korea. For this, our paper sets models and focuses on empirical test. The stock market on this month correlate with industries in Korea. The stock market doesn't lead to industries. The industries and macroeconomic variables have high correlation. We test that gradual diffusion of industrial information will predict stock market in Korea. For this, we analysis on possibility of Granger cause by VAR models between industries and stock market. As a result, 21 portfolios cause to Kospi statistically significance at 5%. Especially, the Beverage portfolio has bilateral Granger causality to Kospi. In case of Internet and Cosmetics portfolio, Kospi has unilateral Granger causality to it. The predictability of specific industries has a relation to Macroeconomic variables. What industrial portfolios predict to Business Coincidence Index? The only 6 industrial portfolios of 36 portfolios have a statistically significance at 10%. And, 9 portfolios have a statistically significance at 5%.

• Journal of Distribution Research
• /
• v.14 no.1
• /
• pp.45-65
• /
• 2009

As a number of people using the internet for their shopping steadily rises, it is increasingly important for retailers to understand why consumers decide to buy products via online or offline. The main purpose of this study is to develop and test a model that enhance our understanding of how consumers respond future online and offline channels for their purchasing. Rather than merely adopting statistical models like most other studies in this field, the present study develops a model that combines double-cohort method with multinomial logit model. It is desirable if one can adopt an overall encompassing criterion in the study of consumer behaviors form diverse sales channels. This study uses the concept of cohort or aging to enable this comparison. It enables us to analyze how consumers respond to online and offline channels as people aged by measuring their shopping behavior for an online and offline retailers and their subsequent purchase intentions. Based on some empirical findings, this study concludes with policy implications and some necessary fields of future studies desirable.

• Proceedings of the Korean Society of Crop Science Conference
• /
• 2017.06a
• /
• pp.303-303
• /
• 2017

Japanese agriculture has faced to several threats: aging and decrease of farmer population, global competition, and the risk of climate change as well as harsh and variable weather. On the other hands, the number of large scale farms is increasing, because farm lands have been being aggregated to fewer numbers of farms. Cost cutting, development of efficient ways to manage complicatedly scattered farm lands, maintaining yield and quality under variable weather conditions, are required to adapt to changing environments. Information and communications technology (ICT) would contribute to solve such problems and to create innovative technologies. Thus we have been developing an early warning and decision support system to reduce weather and climate risks for rice, wheat and soybean production in Japan. The concept and prototype of the system will be shown. The system consists of a weather data system (Agro-Meteorological Grid Square Data System, AMGSDS), decision support contents where information is automatically created by crop models and delivers information to users via internet. AMGSDS combines JMA's Automated Meteorological Data Acquisition System (AMeDAS) data, numerical weather forecast data and normal values, for all of Japan with about 1km Grid Square throughout years. Our climate-smart system provides information on the prediction of crop phenology, created with weather forecast data and crop phenology models, as an important function. The system also makes recommendations for crop management, such as nitrogen-topdressing, suitable harvest time, water control, pesticide spray. We are also developing methods to perform risk analysis on weather-related damage to crop production. For example, we have developed an algorism to determine the best transplanting date in rice under a given environment, using the results of multi-year simulation, in order to answer the question "when is the best transplanting date to minimize yield loss, to avoid low temperature damage and to avoid high temperature damage?".