• Journal of Intelligence and Information Systems
• /
• v.20 no.3
• /
• pp.77-92
• /
• 2014

Recently, numerous documents including unstructured data and text have been created due to the rapid increase in the usage of social media and the Internet. Each document is usually provided with a specific category for the convenience of the users. In the past, the categorization was performed manually. However, in the case of manual categorization, not only can the accuracy of the categorization be not guaranteed but the categorization also requires a large amount of time and huge costs. Many studies have been conducted towards the automatic creation of categories to solve the limitations of manual categorization. Unfortunately, most of these methods cannot be applied to categorizing complex documents with multiple topics because the methods work by assuming that one document can be categorized into one category only. In order to overcome this limitation, some studies have attempted to categorize each document into multiple categories. However, they are also limited in that their learning process involves training using a multi-categorized document set. These methods therefore cannot be applied to multi-categorization of most documents unless multi-categorized training sets are provided. To overcome the limitation of the requirement of a multi-categorized training set by traditional multi-categorization algorithms, we propose a new methodology that can extend a category of a single-categorized document to multiple categorizes by analyzing relationships among categories, topics, and documents. First, we attempt to find the relationship between documents and topics by using the result of topic analysis for single-categorized documents. Second, we construct a correspondence table between topics and categories by investigating the relationship between them. Finally, we calculate the matching scores for each document to multiple categories. The results imply that a document can be classified into a certain category if and only if the matching score is higher than the predefined threshold. For example, we can classify a certain document into three categories that have larger matching scores than the predefined threshold. The main contribution of our study is that our methodology can improve the applicability of traditional multi-category classifiers by generating multi-categorized documents from single-categorized documents. Additionally, we propose a module for verifying the accuracy of the proposed methodology. For performance evaluation, we performed intensive experiments with news articles. News articles are clearly categorized based on the theme, whereas the use of vulgar language and slang is smaller than other usual text document. We collected news articles from July 2012 to June 2013. The articles exhibit large variations in terms of the number of types of categories. This is because readers have different levels of interest in each category. Additionally, the result is also attributed to the differences in the frequency of the events in each category. In order to minimize the distortion of the result from the number of articles in different categories, we extracted 3,000 articles equally from each of the eight categories. Therefore, the total number of articles used in our experiments was 24,000. The eight categories were "IT Science," "Economy," "Society," "Life and Culture," "World," "Sports," "Entertainment," and "Politics." By using the news articles that we collected, we calculated the document/category correspondence scores by utilizing topic/category and document/topics correspondence scores. The document/category correspondence score can be said to indicate the degree of correspondence of each document to a certain category. As a result, we could present two additional categories for each of the 23,089 documents. Precision, recall, and F-score were revealed to be 0.605, 0.629, and 0.617 respectively when only the top 1 predicted category was evaluated, whereas they were revealed to be 0.838, 0.290, and 0.431 when the top 1 - 3 predicted categories were considered. It was very interesting to find a large variation between the scores of the eight categories on precision, recall, and F-score.

• Journal of Intelligence and Information Systems
• /
• v.28 no.1
• /
• pp.131-154
• /
• 2022

Recently, researches on unstructured data analysis have been actively conducted with the development of information and communication technology. In particular, topic modeling is a representative technique for discovering core topics from massive text data. In the early stages of topic modeling, most studies focused only on topic discovery. As the topic modeling field matured, studies on the change of the topic according to the change of time began to be carried out. Accordingly, interest in dynamic topic modeling that handle changes in keywords constituting the topic is also increasing. Dynamic topic modeling identifies major topics from the data of the initial period and manages the change and flow of topics in a way that utilizes topic information of the previous period to derive further topics in subsequent periods. However, it is very difficult to understand and interpret the results of dynamic topic modeling. The results of traditional dynamic topic modeling simply reveal changes in keywords and their rankings. However, this information is insufficient to represent how the meaning of the topic has changed. Therefore, in this study, we propose a method to visualize topics by period by reflecting the meaning of keywords in each topic. In addition, we propose a method that can intuitively interpret changes in topics and relationships between or among topics. The detailed method of visualizing topics by period is as follows. In the first step, dynamic topic modeling is implemented to derive the top keywords of each period and their weight from text data. In the second step, we derive vectors of top keywords of each topic from the pre-trained word embedding model. Then, we perform dimension reduction for the extracted vectors. Then, we formulate a semantic vector of each topic by calculating weight sum of keywords in each vector using topic weight of each keyword. In the third step, we visualize the semantic vector of each topic using matplotlib, and analyze the relationship between or among the topics based on the visualized result. The change of topic can be interpreted in the following manners. From the result of dynamic topic modeling, we identify rising top 5 keywords and descending top 5 keywords for each period to show the change of the topic. Existing many topic visualization studies usually visualize keywords of each topic, but our approach proposed in this study differs from previous studies in that it attempts to visualize each topic itself. To evaluate the practical applicability of the proposed methodology, we performed an experiment on 1,847 abstracts of artificial intelligence-related papers. The experiment was performed by dividing abstracts of artificial intelligence-related papers into three periods (2016-2017, 2018-2019, 2020-2021). We selected seven topics based on the consistency score, and utilized the pre-trained word embedding model of Word2vec trained with 'Wikipedia', an Internet encyclopedia. Based on the proposed methodology, we generated a semantic vector for each topic. Through this, by reflecting the meaning of keywords, we visualized and interpreted the themes by period. Through these experiments, we confirmed that the rising and descending of the topic weight of a keyword can be usefully used to interpret the semantic change of the corresponding topic and to grasp the relationship among topics. In this study, to overcome the limitations of dynamic topic modeling results, we used word embedding and dimension reduction techniques to visualize topics by era. The results of this study are meaningful in that they broadened the scope of topic understanding through the visualization of dynamic topic modeling results. In addition, the academic contribution can be acknowledged in that it laid the foundation for follow-up studies using various word embeddings and dimensionality reduction techniques to improve the performance of the proposed methodology.

• Atmosphere
• /
• v.21 no.4
• /
• pp.429-438
• /
• 2011

The climatology of surface UV radiation for Seoul, presented in Cho et al. (1998; 2001), has been updated using measurement of surface erythemal ultraviolet (EUV) and total ultraviolet (TUV) irradiance (wavelength 286.5~363.0 nm) by a Brewer Spectrophotometer (MK-IV) for the period 2004~2010. The analysis was also carried out together with the broadband total (global) solar irradiance (TR ; 305~2800 nm) and cloud amount to compare with the UV variations, measured by Seoul meteorological station of Korean Meteorological Agency located near the present study site. Under all-sky conditions, the day-to-day variability of EUV exhibits annual mean of 98% in increase and 31% in decrease. It has been also shown that the EUV variability is 17 times as high as the total ozone in positive change, whereas this is 6 times higher in negative change. Thus, the day to day variability is dominantly caused rather by the daily synoptic situations than by the ozone variability. Annual mean value of daily EUV and TUV shows $1.62kJm^{-2}$ and $0.63MJm^{-2}$ respectively, whereas mean value of TR is $12.4MJm^{-2}$ ($143.1Wm^{-2}$). The yearly maximum in noon-time UV Index (UVI) varies between 9 and 11 depending on time of year. The highest UVI shows 11 on 20 July, 2008 during the period 2004~2010, but for the period 1994~2000, the index of 12 was recorded on 13 July, 1994 (Cho et al., 2001). A 40% of daily maximum UVI belongs to "low (UVI < 2)", whereas the UVI less than 5% of the maximum show "very high (8 < UVI < 10)". On average, the maximum UVI exceeded 8 on 9 days per year. The values of Tropospheric Emission Monitoring Internet Service (TEMIS) EUV and UVI under cloud-free conditions are 1.8 times and 1.5 times, respectively, higher than the all-sky measurements by the Brewer. The trend analysis in fractional deviation of monthly UV from the reference value shows a decrease of -0.83% and -0.90% $decade^{-1}$ in the EUV and TUV, respectively, whereas the TR trend is near zero (+0.11% $decade^{-1}$). The trend is statistically significant except for TR trend (p = 0.279). It is possible that the recent UV decrease is mainly associated with increase in total ozone, but the trend in TR can be attributed to the other parameters such as clouds except the ozone. Certainly, the cloud effects suggest that the reason for the differences between UV and TR trends can be explained. In order to estimate cloud effects, the EUV, TUV and TR irradiances have been also evaluated for clear skies (cloud cover < 25%) and cloudy skies (cloud cover ${\geq}$ 75%). Annual mean values show that EUV, TUV and TR are $2.15kJm^{-2}$, $0.83MJm^{-2}$, and $17.9MJm^{-2}$ for clear skies, and $1.24kJm^{-2}$, $0.46MJm^{-2}$, and $7.2MJm^{-2}$ for cloudy skies, respectively. As results, the transmission of radiation through clouds under cloudy-sky conditions is observed to be 58%, 55% and 40% for EUV, TUV and TR, respectively. Consequently, it is clear that the cloud effects on EUV and TUV are 18% and 15%, respectively lower than the effects on TR under cloudy-sky conditions. Clouds under all-sky conditions (average of cloud cover is 5 tenths) reduced the EUV and TUV to about 25% of the clear-sky (cloud cover < 25%) values, whereas for TR, this was 31%. As a result, it is noted that the UV radiation is attenuated less than TR by clouds under all weather conditions.

• Journal of the Korean Institute of Landscape Architecture
• /
• v.46 no.2
• /
• pp.27-36
• /
• 2018

This study suggests the necessity of behavior analysis as changes to a park environment to reflect user desires can be implemented only by grasping the needs of park users. Online data (blog) were defined as the basic data of the study. After collecting data by 5 - year units, data mining was used to derive the characteristics of the time series behavior while the significance of the online data was verified through social network analysis. The results of the text mining analysis are as follows. First, primary results included 'walking', 'photography', 'riding bicycles'(inline, kickboard, etc.), and 'eating'. Second, in the early days of the collected data, active physical activity such as exercise was the main factor, but recent passive behavior such as eating, using a mobile phone, games, food and drinking coffee also appeared as a new behavior characteristic in parks. Third, the factors affecting the behavior of park users are the changes of various conditions of society such as internet development and a culture of expressing unique personalities and styles. Fourth, the special behaviors appearing at Olympic Park were derived from educational activities such as cultural activities including watching performances and history lessons. In conclusion, it has been shown that people's lifestyle changes and the behavior of a park are influenced by the changes of the various times rather than the original purpose that was intended during park planning and design. Therefore, it is necessary to create an environment tailored to users by considering the main behaviors and influencing factors of Olympic Park. Text mining used as an analytical method has the merit that past data can be collected. Therefore, it is possible to form analysis from a long-term viewpoint of behavior analysis as well as to measure new behavior and value with derived keywords. In addition, the validity of online data was verified through social network analysis to increase the legitimacy of research results. Research on more comprehensive behavior analysis should be carried out by diversifying the types of data collected later, and various methods for verifying the accuracy and reliability of large-volume data will be needed.

• Journal of the Korean Academy of Child and Adolescent Psychiatry
• /
• v.17 no.1
• /
• pp.19-26
• /
• 2006

Objectives : The object of this study was to examine computer game playing patterns and psychopathologies related to computer game addiction in school-age children. Methods : The subjects were 533 elementary school students (4th to 6th grade) in Kangdonggu, Seoul. We evaluated computer playing patterns of all subjects using computer game playing pattern questionnaire, and determined the risk group of computer game addiction by internet game addiction scale score. We evaluated subscale score of K-CBCL from parents of all subjects, and conducted correlation analysis and logistic regression analysis between computer game addiction and subscale score of K-CBCL. Results : In 488 responders, 10.2% of started playing computer game in preschool age, and 67.2% started at low grade of elementary school. The mean frequency of computer game play per week was 3.66 days. Mean time spent playing computer games per day was 1.89 hours. 'Simply for fun' was the most common reason far playing computer games (40.8%). Male subjects showed statistically significant differences in age of starting computer game, frequency of computer game play per week, reasons for playing computer game and computer game addiction scale scores. There were significant correlations between computer game addiction scale scores and academic performance, somatic complaints, attention problems, and internalizing problems in K-CBCL. But In logistic regression analysis, only attention problems among K-CBCL subscales showed significant predictability to computer game addiction. Conclusion : Upper grade elementary school students experienced computer game playing at the very early age, and spend much time in playing computer games. There were significant correlation and predictability between computer game addiction and attention problems.

• Journal of the Korean Society of Food Science and Nutrition
• /
• v.36 no.6
• /
• pp.727-734
• /
• 2007

The purpose of this study was to evaluate nutrition knowledge and utilization of information on fast food of secondary school students in the two biggest cities, Seoul and Busan. Questionnaires were distributed to a total of 1,265 students (645 at 10 middle schools, 620 at 10 high schools). An average rate of correct answers on nutrition knowledge of fast food was 73.4%. General nutrition knowledge on fast food showed high percentage of correct answers, but the knowledge of vitamin and fiber on fast food was of low percentage. The scores of nutrition knowledge, the fast food (ie. chicken, pizza, gimbab, dukbokki, and ramyon) intake group with $'{\geq}1$ time/week' showed significantly (p<0.01) lower scores than those of the group with $'{\leq}1$ time/month'. The groups with higher interest in health and weight control had significantly (p<0.05) higher scores on nutrition knowledge than those of lower interest groups. Only 19.6% of the subjects had the experience of using the nutrition information for selecting fast food, but 66% of them wanted the developments of the internet-sites to provide information on fast food and on the relationship between fast food and health.

• Cartoon and Animation Studies
• /
• s.22
• /
• pp.81-97
• /
• 2011

The technology of 'Virtual Reality' has placed in advanced tools for human beings' joy and anger together with sorrow and pleasure in our generation. It has recently tried in a variety ways to use as an implication for treatment in the field of Cognitive Psychology. Especially, it widely approaches to human in terms of that a sense of reality in a virtual world through the five senses should reinterpret the meaning of cognition in the real world. Based on this paradigm shift, it allows for new treatment using the technology of virtual reality. A typical example is a field of Therapy in order to overcome panic disorder. It has advantages that in particular development of flexible interaction technologies in a virtual space can lead patients to experience psychological environments rather than physical one. the interaction technology provides environments in which users' five senses can be actively stimulated, it is very useful that information from the experiences in the virtual world allows people to learn through real experiences by renewing potential energies, advantages of Virtual Reality Therapy can be customized treatment by depending on symptoms in patients with panic disorder and are capable of differentiate application for the cure at each stage. It is to treat by leading patients to get accustomed to environments and situations in real world through care process with each symptom and stage. It is helpful that based on A Human-Sensibility Ergonomics, technologies like immersive virtual reality equipment, force-relative feedback and stereophonic sound, and like stimulating the sense of smell make people to induce experiences by stimulating human's five senses. There are many advantages of immersion in virtual world in that the phenomenon such as challenge, interaction, reality, illusion, and cooperation is expanded. As an application for therapy by growing such augmented reality, virtual space and sharing of data through the Internet and also inexpensive its availability have recently expanded the base. There are other benefits of Virtual Reality Therapy offering active interaction environments for cognitive experience which can provide appropriately adjusted environments for patients who are hard to overcome the real situation because of phobia. In addition to that it is safe and economical and patients' confidentiality is assured. Moreover, due to the principles of applying real-time navigation the Virtual Reality Therapy makes modification and supplementation easier and also it can reduce cybersickness because of the supply of Lenticular allowing people to see stereoscopy without eyeglasses, which makes sense of presence clearer. On top of that due to the development of interactive technologies, it is becoming close to sense of reality similar to real world by leading users to navigate by themselves and to operate objects in a virtual space. This paper will therefore examine, although it is of limited, characteristics of application of virtual reality technology based on A Human-Sensibility Ergonomics used for treatment for a disorder. this paper will analyse a range of its application and problems and it will suggest the future possibilities.

• Journal of dental hygiene science
• /
• v.17 no.6
• /
• pp.501-507
• /
• 2017

The shortage of dental hygienists has been a long-standing problem in Korea. Small-scaled dental clinics suffer from a lack of dental hygienists, who seem to prefer working at large-scaled dental clinics. The purpose of this study was to confirm the differences in the working conditions according to the scales of dental clinics. We collected the working information registered via job advertisements through the web-sites of Korean Dental Hygienists Association, Dental Jobs, and Nurse Jobs from July to August 2016. The results were as follows: 96.7% of the advertisements wanted regular workers, while the proportion of part-time workers was the highest (34.8%) in the group with less than 3 employees. The average workdays per week was $5.32{\pm}0.55$ days, and the group with less than 3 employees had significantly longer workdays than the other groups. The daily working time was $8.99{\pm}0.44$ hours, and there was no difference among the groups. Night overtime hours were needed by 54.4%, 45.0%, and 31.3% of the groups with of the groups with 4~7 employees, more than 8 employees, and less than 3 employees, respectively. Information regarding annual leave (60.5%), monthly leave (63.9%), half a day off (32.4%) and vacations (43.1%) were presented in the job advertisements, and these proportions were significantly higher by the group with more than 8 employees. Information on overtime pay (14.4%), night-work pay (13.4%), incentives (34.1%), lunches (60.2%), vacation bonuses (33.8%), and self-development (20.4%) were presented in job advertisements. The group with 4~7 employees had significantly higher proportions in severance pay, vacation bonuses, self-development, and major national insurance. It is necessary to consider the improvement of working conditions, diversity of working styles, and welfare of dental hygienists, and it is suggested that small dental clinics provide more precise working conditions.

• Journal of Intelligence and Information Systems
• /
• v.18 no.3
• /
• pp.97-117
• /
• 2012

Nowadays, Internet users tend to do a variety of actions at the same time such as web browsing, social networking and multimedia consumption. While watching a video, once a user is interested in any product, the user has to do information searches to get to know more about the product. With a conventional approach, user has to search it separately with search engines like Bing or Google, which might be inconvenient and time-consuming. For this reason, a video annotation platform has been developed in order to provide users more convenient and more interactive ways with video content. In the future of smart TV environment, users can follow annotated information, for example, a link to a vendor to buy the product of interest. It is even better to enable users to search for information by directly discussing with friends. Users can effectively get useful and relevant information about the product from friends who share common interests or might have experienced it before, which is more reliable than the results from search engines. Social networking services provide an appropriate environment for people to share products so that they can show new things to their friends and to share their personal experiences on any specific product. Meanwhile, they can also absorb the most relevant information about the product that they are interested in by either comments or discussion amongst friends. However, within a very huge graph of friends, determining the most appropriate persons to ask for information about a specific product has still a limitation within the existing conventional approach. Once users want to share or discuss a product, they simply share it to all friends as new feeds. This means a newly posted article is blindly spread to all friends without considering their background interests or knowledge. In this way, the number of responses back will be huge. Users cannot easily absorb the relevant and useful responses from friends, since they are from various fields of interest and knowledge. In order to overcome this limitation, we propose a method to filter a user's friends for information search, which leverages semantic video annotation and social networking services. Our method filters and brings out who can give user useful information about a specific product. By examining the existing Facebook information regarding users and their social graph, we construct a user profile of product interest. With user's permission and authentication, user's particular activities are enriched with the domain-specific ontology such as GoodRelations and BestBuy Data sources. Besides, we assume that the object in the video is already annotated using Linked Data. Thus, the detail information of the product that user would like to ask for more information is retrieved via product URI. Our system calculates the similarities among them in order to identify the most suitable friends for seeking information about the mentioned product. The system filters a user's friends according to their score which tells the order of whom can highly likely give the user useful information about a specific product of interest. We have conducted an experiment with a group of respondents in order to verify and evaluate our system. First, the user profile accuracy evaluation is conducted to demonstrate how much our system constructed user profile of product interest represents user's interest correctly. Then, the evaluation on filtering method is made by inspecting the ranked results with human judgment. The results show that our method works effectively and efficiently in filtering. Our system fulfills user needs by supporting user to select appropriate friends for seeking useful information about a specific product that user is curious about. As a result, it helps to influence and convince user in purchase decisions.

• The Journal of Natural Sciences
• /
• v.10 no.1
• /
• pp.209-216
• /
• 1998

The results of the study of diving the 21st century's Korea fashion market into consumer market, fashion market, and a new marketing strategy are as follows. The 21st consumer market is First, a fashion democracy phenomenon. As many people try to leave unconditional fashion following, consumer show a phenomenon to choose and create their own fashion by subjective judgements. Second, a phenomenon of total fashion pursuit. Consumer in the future are likely to put their goals not in differentiating small item products, but considering various fashion elements based on their individuality and sense of value. Third, world quality-oriented. With the improvement of life level, it accomplishes to emphasize consumers' fashion mind on the world wide popular use of materials, quality, design and brand image. Fourth, with the entrance of neo-rationalism, consumers show increasing trends to emphasize wisdom, solidity in goods strategy pursuing high quality fashion and to demand resonable prices. Fifth, concept-oriented. Consumers are changing into pursuing concept appropriate to individual life scene. Prospecting the composition of the 21st century's fashion market, First, sportive casual zone will draw attention more than any other zone. This is because interest in sports will grow according to the increase of leisure time and the expasion of time and space in the 21st century, and also ecology will become the important issue of sports sense because of human beings's natural habit toward nature. Second, the down aging phenomenon will accelerate its speed as a big trend. Third, a retro phenomenon, a concept contrary to digital and high-tech, will become another big trend for its remake, antique, and classic concept in fashion market with ecology trend. New marketing strategy to cope with changing fashion market is as follows. First, with the trend of borderless concept, borders between apparels are becoming vague, for example, they offer custom-made products to consumers. Second, as more enterprises take the way of gorilla and guerrilla where guerrillas who aim at niche market show up will develop. Basically, they think highly of individual creative study, and pursue the scene adherence with high sensitiveness. However this polarization becomes mutually-supplementing relationship showing gorilla's guerilla movement, and guerilla's gorilla high-tech. Third with the development of value retailing, enterprises pursuing mass merchandising of groups called category killers are expanded and amplified to new product fields, and expand business' share. Fourth, using outsourcing, the trend to use exterior function leaving each enterprise's strength by inspecting its own work is gradually strong. Fifth, with the expansion of none store sale, the entrance of the internet and the CD-ROM sales added to communication sales such as catalogues are specified. An eminent American think tank expect that 5-5% of the total sale of clothes and home goods in 2010 will be done by none store sale. Accordingly, to overcome the problems, First international, global level marketing, Second, the improvement of technology, Third, knowledge-creating marketing are needed.