• Proceedings of the KAIS Fall Conference
• /
• 2003.11a
• /
• pp.141-148
• /
• 2003

Customer retention is a common concern for many industries and a critical issue for the survival in today's greatly compressed marketplace. Current customer retention models only focus on detection of potential defectors based on the likelihood of defection by using demographic and customer profile information. In this paper, we propose a dynamic procedure for defection detection and prevention using past and current customer behavior by utilizing SOM and Markov chain. The basic idea originates from the observation that a customer has a tendency to change his behavior (i.e. trim-out his usage volumes) before his eventual withdrawal. This gradual pulling out process offers the company the opportunity to detect the defection signals. With this approach, we have two significant benefits compared with existing defection detection studies. First, our procedure can predict when the potential defectors could withdraw and this feature helps to give marketing managers ample lead-time for preparing defection prevention plans. The second benefit is that our approach can provide a procedure for not only defection detection but also defection prevention, which could suggest the desirable behavior state for the next period so as to lower the likelihood of defection. We applied our dynamic procedure for defection detection and prevention to the online gaming industry. Our suggested procedure could predict potential defectors without deterioration of prediction accuracy compared to that of the MLP neural network and DT.

• Journal of the Korean Society for Library and Information Science
• /
• v.37 no.2
• /
• pp.153-168
• /
• 2003

A hyperlink on academic World Wide Web has started to be recognized as a form of collaborative communication network connecting individual researchers and research groups and expanding their collaboration relations by making possible easy and direct online contact among people or groups anywhere in the world. This paper describes the structure of academic hyperlinks embedded in universities' Web sites hosted at the 10 Asian countries and further, examines the association between the structure of the hyperlink network and collaborative communication pattern among those countries based on their frequency of co-authoring articles. This research found that the number of inter-hyperlinks among universities' Web sites was significantly correlated with the frequency of co-authored articles across the 10 countries.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.2
• /
• pp.421-441
• /
• 2021

Fruit detection in orchards is one of the most crucial tasks for designing the visual system of an automated harvesting robot. It is the first and foremost tool employed for tasks such as sorting, grading, harvesting, disease control, and yield estimation, etc. Efficient visual systems are crucial for designing an automated robot. However, conventional fruit detection methods always a trade-off with accuracy, real-time response, and extensibility. Therefore, an improved method is proposed based on coarse-to-fine multitask cascaded convolutional networks (MTCNN) with three aspects to enable the practical application. First, the architecture of Fruit-MTCNN was improved to increase its power to discriminate between objects and their backgrounds. Then, with a few manual labels and operations, synthetic images and labels were generated to increase the diversity and the number of image samples. Further, through the online hard example mining (OHEM) strategy during training, the detector retrained hard examples. Finally, the improved detector was tested for its performance that proved superior in predicted accuracy and retaining good performances on portability with the low time cost. Based on performance, it was concluded that the detector could be applied practically in the actual orchard environment.

• The Journal of Asian Finance, Economics and Business
• /
• v.8 no.3
• /
• pp.61-67
• /
• 2021

The covid-19 pandemic scenario caused the most extensive economic shocks the world has experienced in decades. Maintaining financial performance and economic stability is essential during the pandemic period. In these conditions, where movement is severely restricted, media consumption is considered to be increasing. The social media platform is one of the media online used by the public as a source of information and also expressing their sentiment, including individual investors in the capital market as social media users. Twitter is one of the social media microblogging platforms used by individual investors to share their opinion and get information. This study aims to determine whether microblogging sentiment investors can predict the capital market during pandemics. To analyze microblogging sentiment investors, we classified sentiment using the phyton text mining algorithm and Naïve Bayesian text classification into level positive, negative, and neutral from November 2019 to November 2020. This study was on 68 listed companies on the Indonesia stock exchange. A Vector Autoregression and Impulse Response is applied to capture short and long-term impacts along with a causal relationship. We found that microblogging sentiment investor has a significant impact on stock returns and volatility and vice-versa. Also, the response due to shocks is convergent, and microblogging investors in Indonesia are categorized as a "news-watcher" investor.

• Journal of the Korean Society for Library and Information Science
• /
• v.55 no.3
• /
• pp.397-417
• /
• 2021

As review text contains the experience and opinions of the customers, analyzing review text helps to understand the subject. Existing studies either only used sentiment analysis on online restaurant reviews to identify the customers' assessment on different features of the restaurant or network analysis to figure out the customers' preference. In this study, we conducted both sentiment analysis and network analysis on the review text of the restaurants with high star ratings and those with low star ratings. We compared the review text of the two groups to distinguish the difference of the two and identify what makes great restaurants great.

• International Journal of Internet, Broadcasting and Communication
• /
• v.14 no.3
• /
• pp.85-90
• /
• 2022

In this study, keywords from representative online portal sites such as NAVER, Google, and Youtube were collected based on text mining analysis technique using TEXTOM to check the changes in the restaurant industry before and after COVID-19. The collection keywords were selected as dining out, food service industry, and dining out culture. For the collected data, the top 30 words were derived, respectively, through the refinement process. In addition, comparative analysis was conducted by defining data from 2018 to 2019 before COVID-19, and from 2020 to 2021 after COVID-19. As a result, 8272 keywords before COVID-19 and 9654 keywords after COVID-19, a total of 17926 keywords, were derived. In order for the food service industry to develop after the COVID-19 pandemic, it is necessary to commercialize the recipes of restaurants to revitalize the distribution of home-use food products that replace home-cooked meals such as meal kits. Due to the social distancing caused by COVID-19, the dining out culture has changed and the trend has changed, and it has been confirmed that the consumption culture has changed to eating and delivering at home more safely than visiting restaurants. In addition, it has been confirmed that the consumption culture of existing consumers is changing to a trend of cooking at home rather than visiting restaurants.

• 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 the Korean Society of Clothing and Textiles
• /
• v.46 no.6
• /
• pp.1004-1022
• /
• 2022

This study seeks to investigate K-fashion's external image by examining the trends in global media reporting. It applies Dynamic Topic Modeling (DTM), which captures the evolution of topics in a sequentially organized corpus of documents, and consists of text preprocessing, the determination of the number of topics, and a timeseries analysis of the probability distribution of words within topics. The data set comprised 551 online media articles on 'Korean fashion' or 'K-fashion' published on Google News between 2010 and 2021. The analysis identifies seven topics: 'brand look and style,' 'lifestyle,' 'traditional style,' 'Seoul Fashion Week (SFW) event,' 'model size,' 'K-pop,' and 'fashion market,' as well as annual topic proportion trends. It also explores annual word changes within the topic and indicates increasing and decreasing word patterns. In most topics, the probability distribution of the word 'brand' is confirmed to be on the increase, while 'digital,' 'platform,' and 'virtual' have been newly created in the 'SFW event' topic. Moreover, this study confirms the transition of each K-fashion topic over the past 12 years, along with various factors related to Hallyu content, traditional culture, government support, and digital technology innovation.

• Safety and Health at Work
• /
• v.14 no.1
• /
• pp.10-16
• /
• 2023

Whilst Ecological momentary assessment (EMA) can provide important insights over time and across contexts among rotation workers whose work periods alternate with leave at home, it can also be challenging to implement in the resources and construction sectors. This review aimed to provide a summary of the methodological characteristics of EMA studies assessing health outcomes and related behaviors in rotation workers. Systematic searches in PubMed, Medline, EMBASE, CINAHL, PsycINFO, and Scopus were done to include 23 studies using EMA methods in assessing health-related outcomes and behaviors. EMA designs included daily diary: assessments once per day typically fixed at the end of day (47.8%), within day fixed interval time-based design: assessments on multiple times per day at certain times of day (17.4%) and combination of both designs (34.8%). Studies employed paper and pencil diaries (73.9%) and one or more electronic methods (60.9%): wrist-worn actigraphy device (52.2%) and online-based diaries (26.1%) for data collection. Most of the studies (91.3%) did not report prompting -EMAs by schedule alerts or compliance. Daily diary and within day fixed interval dairies designs are common, with the increasing use of electronic EMA delivery techniques. It is unclear how well participants adhere to assessment schedules, as these are inadequately reported. Researchers should report compliance-related information.

• Annual Conference of KIPS
• /
• 2020.05a
• /
• pp.389-392
• /
• 2020

The growth in social media, blogs and restaurant listing directories have led to increasing customer reviews about restaurants, their quality of food items and services available on the internet. These user reviews offer a massive amount of valuable information that can be used for various decision-making purposes. Currently, most food recommendation sites provide recommendation scores about restaurants rather than food items of the restaurant and the provided recommendation scores may be biased since they are calculated only from user reviews listed only in their sites. Usually, people wants a reliable recommendation about foods, not restaurant. In this paper, we present a reliable Korean food items rating method; we first extract food items by applying NER technique to restaurant reviews collected from many Korean restaurant recommendation web sites, blogs and web data. Then, we apply lexicon-based sentiment analysis on collected user reviews and predict people's opinions as sentiment polarity scores (+1 for positive; -1 for negative; 0 for neutral). Finally, by taking average of all calculated polarity scores about a food item, we obtain a rating to individual menu items of the restaurant. The proposed food item rating is more reliable since it does not depend on reviews of only one site.