• Radiation Oncology Journal
• /
• v.23 no.3
• /
• pp.143-156
• /
• 2005

Background: The best dose-fractionation regimen of the definitive radiotherapy for cervix cancer remains to be clearly determined. It seems to be partially attributed to the complexity of the affecting factors and the lack of detailed information on external and intra-cavitary fractionation. To find optimal practice guidelines, our experiences of the combination of external beam radiotherapy (EBRT) and high-dose-rate intracavitary brachytherapy (HDR-ICBT) were reviewed with detailed information of the various treatment parameters obtained from a large cohort of women treated homogeneously at a single institute. Materials and Methods: The subjects were 743 cervical cancer patients (Stage IB 198, IIA 77, IIB 364, IIIA 7, IIIB 89 and IVA 8) treated by radiotherapy alone, between 1990 and 1996. A total external beam radiotherapy (EBRT) dose of $23.4\~59.4$ Gy (Median 45.0) was delivered to the whole pelvis. High-dose-rate intracavitary brachytherapy (HDR-IBT) was also peformed using various fractionation schemes. A Midline block (MLB) was initiated after the delivery of $14.4\~43.2$ Gy (Median 36.0) of EBRT in 495 patients, while In the other 248 patients EBRT could not be used due to slow tumor regression or the huge initial bulk of tumor. The point A, actual bladder & rectal doses were individually assessed in all patients. The biologically effective dose (BED) to the tumor ($\alpha/\beta$=10) and late-responding tissues ($\alpha/\beta$=3) for both EBRT and HDR-ICBT were calculated. The total BED values to point A, the actual bladder and rectal reference points were the summation of the EBRT and HDR-ICBT. In addition to all the details on dose-fractionation, the other factors (i.e. the overall treatment time, physicians preference) that can affect the schedule of the definitive radiotherapy were also thoroughly analyzed. The association between MD-BED $Gy_3$ and the risk of complication was assessed using serial multiple logistic regression models. The associations between R-BED $Gy_3$ and rectal complications and between V-BED $Gy_3$ and bladder complications were assessed using multiple logistic regression models after adjustment for age, stage, tumor size and treatment duration. Serial Coxs proportional hazard regression models were used to estimate the relative risks of recurrence due to MD-BED $Gy_{10}$, and the treatment duration. Results: The overall complication rate for RTOG Grades $1\~4$ toxicities was $33.1\%$. The 5-year actuarial pelvic control rate for ail 743 patients was $83\%$. The midline cumulative BED dose, which is the sum of external midline BED and HDR-ICBT point A BED, ranged from 62.0 to 121.9 $Gy_{10}$ (median 93.0) for tumors and from 93.6 to 187.3 $Gy_3$ (median 137.6) for late responding tissues. The median cumulative values of actual rectal (R-BED $Gy_3$) and bladder Point BED (V-BED $Gy_3$) were 118.7 $Gy_3$ (range $48.8\~265.2$) and 126.1 $Gy_3$ (range: $54.9\~267.5$), respectively. MD-BED $Gy_3$ showed a good correlation with rectal (p=0.003), but not with bladder complications (p=0.095). R-BED $Gy_3$ had a very strong association (p=<0.0001), and was more predictive of rectal complications than A-BED $Gy_3$. B-BED $Gy_3$ also showed significance in the prediction of bladder complications in a trend test (p=0.0298). No statistically significant dose-response relationship for pelvic control was observed. The Sandwich and Continuous techniques, which differ according to when the ICR was inserted during the EBRT and due to the physicians preference, showed no differences in the local control and complication rates; there were also no differences in the 3 vs. 5 Gy fraction size of HDR-ICBT. Conclusion: The main reasons optimal dose-fractionation guidelines are not easily established is due to the absence of a dose-response relationship for tumor control as a result of the high-dose gradient of HDR-ICBT, individual differences In tumor responses to radiation therapy and the complexity of affecting factors. Therefore, in our opinion, there is a necessity for individualized tailored therapy, along with general guidelines, in the definitive radiation treatment for cervix cancer. This study also demonstrated the strong predictive value of actual rectal and bladder reference dosing therefore, vaginal gauze packing might be very Important. To maintain the BED dose to less than the threshold resulting in complication, early midline shielding, the HDR-ICBT total dose and fractional dose reduction should be considered.

• Journal of Nutrition and Health
• /
• v.45 no.6
• /
• pp.562-576
• /
• 2012

Han-sik is a term in Korean that may indicate any Korean dish or food. At present, there is no general consensus on the definition of Han-sik among scholars or professionals in related fields. The aim of this study was to investigate perceptions of Han-sik by professionals in the fields of food, nutrition, and culinary arts using 512 dishes and foods commonly consumed by Koreans using the 4th Korean National Health and Nutrition Survey. A total of 117 professionals out of 185 initially contacted professionals participated in this online survey. We calculated the rate of respondents with a positive answer, that is "It is Han-sik', on each dish and food from the 512 items in 28 dish groups. Items were categorized into five groups according to their Han-sik perception rate: over 90%, 75-89%, 50-74%, 25-49%, and below 25%. Most items in the three dish groups 'Seasoned vegetables, cooked (Namul Suk-chae)', 'Kimchis', and 'Salt-fermented foods (Jeotgal)' showed high perception rates of Han-sik, with a higher than 90% positive response. Items in 'Soups', 'Stews', and 'Steamed foods' dish groups also showed high perception rates of Han-sik. However, no item showed a greater than 90% Han-sik perception rate in 'Fried foods (Twigim)', 'Meat, poultry and fishes', 'Legumes, nuts, and seeds', 'Milk and milk products', 'Sugars and confectioneries', and 'Soup'. Most items in the 'Milk and milk products', 'Sugars and confectioneries', and 'Soup' groups belonged to the lowest perception rate of below 25%. There was a significant difference in the proportion of items perceived as Han-sik by the length of living abroad to (p < 0.05). In summary, the perception rate of Han-sik seemed to be affected by the cooking method, ingredients, and length of time living abroad by the professionals. Further studies targeting subjects with different characteristics and socioeconomic status are warranted to define the concept of Han-sik.

• Journal of Preventive Medicine and Public Health
• /
• v.31 no.4 s.63
• /
• pp.751-769
• /
• 1998

This study was conducted to assess the attitude of pregnant women toward delivery method, understanding of the reason for determining her own delivery method, participation in decision-making process and satisfaction with delivery method after labor. Study subjects were 693 pregnant women who had visited obstetric clinic for prenatal care in the last month of pregnancy in one general hospital and one obstetrics-gynecology specialty hospital in Taegu city from February 1 to March 31 in 1998. A questionnaire was administered before and after labor and a telephone interview was done one month after labor. Proportion of women who had health education and/or counselling about delivery method during prenatal care was 24.0% and this proportion was higher for women who had previous c-section(35.5%) than others. Women thought vaginal delivery is better than c-section for both maternal and baby's health regardless of previous delivery method. About 90% of primipara and multiparous women who had previous vaginal delivery wanted vaginal delivery for the index birth, while 85.6% of multiparous women who had previous c-section wanted repeat c-section. Reasons for choosing c-section in pregnant women who preferred vaginal delivery before labor were recommendation of doctors(81.9%), recommendation of husband (0.8%), agreement between doctor and pregnant woman(4.7%), and mother's demand (12.6%). Reasons for choosing vaginal delivery were mother's demand(30.6%) and no indication for c-section(67.2%). Reasons for choosing c-section in pregnant women who preferred c-section before labor were recommendation of doctors(76.2%), mother's demand(20.0%), recommendation of husband(1.3%), and agreement between doctor and pregnant woman(2.5%). Of the pregnant women who had c-section, by doctor's recommendation, the proportion of women who had heard detailed explanation about reason for c-section by doctor was 55.1%. Mother's statement about the reason for c-section was consistent with the medical record in 75.9% . However, over 5% points disparities were shown between mother's statement and medical record in cases of the repeat c-section and mother's demand. In primipara and multiparous women who had previous vaginal delivery, the delivery method for index birth had statistically significant association with the preference of delivery method before labor(p<0.05). All of the women who had previous c-section had delivered the index baby by c-section. Among mothers who had delivered the index baby vaginally, 84.9% of them were satisfied with their delivery method immediately after labor and 85.1% at 1 month after labor. However, mothers who had c-section stated that they are satisfied with c-section in 44.6% immediately after labor and 42.0% at 1 month after labor. Preferred delivery method for the next birth had statistically significant association with delivery method for the index birth both immediately after labor and in 1 month after labor. The proportion of mothers who prefer vaginal delivery for the next birth increased with the degree of satisfaction with the vaginal delivery for the index birth but the proportion of mothers who prefer c-section for the next birth was high and they did not change significantly with the degree of satisfaction with the c-section for the index birth. These results suggest that the current high technology-based, physician-centered prenatal and partritional cares need to be reoriented to the basic preventive and promotive technology-based, and mother-fetus-centered care. It is also suggested that active involvement of pregnant woman in decision-making process for the delivery method will increase the rate of vaginal birth after c-section and decrease c-section rate and improve the degree of maternal satisfaction after delivery.

• Journal of Global Scholars of Marketing Science
• /
• v.20 no.2
• /
• pp.179-187
• /
• 2010

Understanding consumers' preferences and use of media types is imperative for marketing and advertising managers, especially in today's fragmented market. A clear understanding assists managers in making more effective selections of appropriate media outlets, yet individuals' choices of type and use of media are based on a variety of characteristics. This paper examines one personality trait, sensation seeking, which has not appeared in the literature examining "new" media preferences and use. Sensation seeking is a personality trait defined as "the need for varied, novel, and complex sensations and experiences and the willingness to take physical and social risks for the sake of such experiences" (Zuckerman 1979). Six hypotheses were developed from a review of the literature. Particular attention was given to the Uses and Gratification theory (Katz 1959), which explains various reasons why people choose media types and their motivations for using the different types of media. Current theory suggests that High Sensation Seekers (HSS), due to their needs for novelty, arousal and unconventional content and imagery, would exhibit higher frequency of use of new media. Specifically, we hypothesize that HSS will use the internet more than broadcast (H1a) or print media (H1b) and more than low (LSS) (H2a) or medium sensation seekers (MSS) (H2b). In addition, HSS have been found to be more social and have higher numbers of friends therefore are expected to use social networking websites such as Facebook/MySpace (H3) and chat rooms (H4) more than LSS (a) and MSS (b). Sensation seekers can manifest into a range of behaviors including disinhibition,. It is expected that alternative social networks such as Facebook/MySpace (H5) and chat rooms (H6) will be used more often for those who have higher levels of disinhibition than low (a) or medium (b) levels. Data were collected using an online survey of participants in extreme sports. In order to reach this group, an improved version of a snowball sampling technique, chain-referral method, was used to select respondents for this study. This method was chosen as it is regarded as being effective to reach otherwise hidden population groups (Heckathorn, 1997). A final usable sample of 1108 respondents, which was mainly young (56.36% under 34), male (86.1%) and middle class (58.7% with household incomes over USD 50,000) was consistent with previous studies on sensation seeking. Sensation seeking was captured using an existing measure, the Brief Sensation Seeking Scale (Hoyle et al., 2002). Media usage was captured by measuring the self reported usage of various media types. Results did not support H1a and b. HSS did not show higher levels of usage of alternative media such as the internet showing in fact lower mean levels of usage than all the other types of media. The highest media type used by HSS was print media, suggesting that there is a revolt against the mainstream. Results support H2a and b that HSS are more frequent users of the internet than LSS or MSS. Further analysis revealed that there are significant differences in the use of print media between HSS and LSS, suggesting that HSS may seek out more specialized print publications in their respective extreme sport activity. Hypothesis 3a and b showed that HSS use Facebook/MySpace more frequently than either LSS or MSS. There were no significant differences in the use of chat rooms between LSS and HSS, so as a consequence no support for H4a, although significant for MSS H4b. Respondents with varying levels of disinhibition were expected to have different levels of use of Facebook/MySpace and chat-rooms. There was support for the higher levels of use of Facebook/MySpace for those with high levels of disinhibition than low or medium levels, supporting H5a and b. Similarly there was support for H6b, Those with high levels of disinhibition use chat-rooms significantly more than those with medium levels but not for low levels (H6a). The findings are counterintuitive and give some interesting insights for managers. First, although HSS use online media more frequently than LSS or MSS, this groups use of online media is less than either print or broadcast media. The advertising executive should not place too much emphasis on online media for this important market segment. Second, social media, such as facebook/Myspace and chatrooms should be examined by managers as potential ways to reach this group. Finally, there is some implication for public policy by the higher levels of use of social media by those who are disinhibited. These individuals are more inclined to engage in more socially risky behavior which may have some dire implications, e.g. by internet predators or future employers. There is a limitation in the study in that only those who engage in extreme sports are included. This is by nature a HSS activity. A broader population is therefore needed to test if these results hold.

• Journal of Preventive Medicine and Public Health
• /
• v.24 no.3 s.35
• /
• pp.363-372
• /
• 1991

We used the health screening data of some rural and urban residents to examine the cross-sectional association between leukocyte count and hypertension. The 206 male and 203 female rural residents were selected by multi-stage cluster sampling method in Kyungsan-Kun area of Kyungbuk province in 1985 and 600 urban residents were selected by the same sampling method as the rural residents in Daegu city of the same province in 1986 compatible with age-sex distribution of Daegu city of 1985 census, but of whom 384 actually responded. The rest of 600 were replaced by age and sex with those who were members of the medical insurance plan visiting the health management department of the university hospital to get the biannual preventive medical checkups. Excluded in the analysis were those having hypertensive history, diseases and extreme outlying values of the screening tests, leaving 373 rural and 571 urban residents. Leukocyte count was measured with ELT-8 Laser shadow method and the unit $cells/mm^3$, Blood pressures were determined with an aneroid sphygmomanometer with pre-standardized method and hypertensives were defined as those showing systolic blood pressure more than 140mmHg and/or diastolic blood pressure more than 90mmHg. Total residents pooled (N=944) showed a significant difference between hypertensives and normotensives ($6965.93{\pm}1997.01\;vs\;6490.61{\pm}1941.32,\;P=0.00$) and in rural residents was noted the similar significant difference (P=0.03). None of significant differences were noted in any stratum stratified by residency and sex. Compared to the lowest quintile of WBC, 2/5 quintile showed odds ratio 0.99 (95% Confidence interval, Ci 0.62-1.59), 3/5 quintile 1.41 (95% CI 0.90-2.21), 4/5 quintile 1.76 (95% CI. 1.14-2.72), and highest quintile 1.80 (1.15-2.82) in the total residents. Likelihood ratio test for linear trend for it indicated a significant trend ($X^2_{trend}=5.53,\;df=1,\;P<0.05$). There were no other significant odds ratios compared to the lowest quintile of WBC in strata stratified by residency and sex. The odds ratios in total residents which had showed significant odds ratios became nonsignificant and of reduced magnitude after controlling age, frequency of smoking and drinking with multiple logistic. regression. In each stratum, it changed magnitudes of odds ratios slightly and unstably. None of the trend tests showed any significant trend. These results suggest that the Friedman et al's finding of association between leukocyte count and hypertension may be due to an statistical type I error resulting from the data dredging in an exploratory study, in which more than 800 variables were screened as possible predictors of hypertension.

• Journal of agricultural medicine and community health
• /
• v.25 no.1
• /
• pp.65-83
• /
• 2000

To know the relationship of general characteristics with activities of daily living(ADL) and instrumental activities of daily living(IADL), we carried out the study on the elderies living in Chollabuk-Do area during 6 months, from June to December in 1999. Study subjects were 281, women and men were 195(69.6%) and 85(30.4%) respectively. Mean ages of women and men were 71.9 and 70.8 respectively. 81.1% elderies has disease and 18.9% were disease free. Disease prevalences of movement joint disease, others, circulatory disease, digestive disease, dental disease, respiratory disease were 50.1%, 25.0%, 10.5%, 9.4%, 8.5%, and 6.3% respectively. The percentages to the use of medical institution in recent were 40.0% for hospital, 16.8% for oriental hospital, 14.5% for public health center, 10.9% for drug store, 10.0% for others, and 7.8% for dental service. The percentages to the improvement of symptom after the use of medical institution were 62.3% for normal, 19.4% for improvement, and 18.2% for non-improvement. The percentages to the health situation were 37.1% for bad, 35.7% for good, and 27.1% for normal. Activities of daily living were 67.1% for 6 scores, 27.9% for 5 scores, 2.1% for 4 scores and ADL of women was lower than the men's. Instrumental activities of daily living were 50.4% for 5 scores, 19.3% for 3 scores, 12.1% for 4 scores and IADL of women was lower than the men's. Frequencies of disability in ADL were 28.9% for incontinence, 6.1% for bathing, 2.9% for meal, 2.5% for walking around house, 1.8% for toilet use, 1.4% for dressing and disability frequencies of women in 6 items of ADL were higher than the men's. The percentages of high, intermediate, low ADL in activities of daily living were 67.1%, 32.5%, 0.4% respectively and decrease of high ADL, increase of intermediate ADL were found with the increasing of age. Frequencies of disability in IADL were 42.9% for payment in and out, 31.8% for payment of written claim, 21.1% for shopping, 16.4% for preparation of meal, and 11.8% for use of bus. All items of women in IADL was higher than the men's but preparation of meal. The percentages of high, intermediate, low IADL in instrumental activities of daily living were 50.4%, 42.5%, 7.1% and decrease of high IADL, increase of intermediate IADL were found with the increasing of age. Mean of ADL with the general characteristics was 5.56 and 2 variables of level of education, health situation were statistically significant. Mean of IADL with the general characteristics was 3.76 and 8 variables of age, sex, level of education, occupation, presence of spouse, duty of living cost, health situation, category of ADL were statistically significant. With the result of stepwise regression, ADL was statistically related with religion, health situation and ADL was statistically related with level of education, living together with family, duty of living cost, health situation.

• Journal of Intelligence and Information Systems
• /
• v.19 no.3
• /
• pp.57-71
• /
• 2013

Over a billion people in the world generate new news minute by minute. People forecasts some news but most news are from unexpected events such as natural disasters, accidents, crimes. People spend much time to watch a huge amount of news delivered from many media because they want to understand what is happening now, to predict what might happen in the near future, and to share and discuss on the news. People make better daily decisions through watching and obtaining useful information from news they saw. However, it is difficult that people choose news suitable to them and obtain useful information from the news because there are so many news media such as portal sites, broadcasters, and most news articles consist of gossipy news and breaking news. User interest changes over time and many people have no interest in outdated news. From this fact, applying users' recent interest to personalized news service is also required in news service. It means that personalized news service should dynamically manage user profiles. In this paper, a content-based news recommendation system is proposed to provide the personalized news service. For a personalized service, user's personal information is requisitely required. Social network service is used to extract user information for personalization service. The proposed system constructs dynamic user profile based on recent user information of Facebook, which is one of social network services. User information contains personal information, recent articles, and Facebook Page information. Facebook Pages are used for businesses, organizations and brands to share their contents and connect with people. Facebook users can add Facebook Page to specify their interest in the Page. The proposed system uses this Page information to create user profile, and to match user preferences to news topics. However, some Pages are not directly matched to news topic because Page deals with individual objects and do not provide topic information suitable to news. Freebase, which is a large collaborative database of well-known people, places, things, is used to match Page to news topic by using hierarchy information of its objects. By using recent Page information and articles of Facebook users, the proposed systems can own dynamic user profile. The generated user profile is used to measure user preferences on news. To generate news profile, news category predefined by news media is used and keywords of news articles are extracted after analysis of news contents including title, category, and scripts. TF-IDF technique, which reflects how important a word is to a document in a corpus, is used to identify keywords of each news article. For user profile and news profile, same format is used to efficiently measure similarity between user preferences and news. The proposed system calculates all similarity values between user profiles and news profiles. Existing methods of similarity calculation in vector space model do not cover synonym, hypernym and hyponym because they only handle given words in vector space model. The proposed system applies WordNet to similarity calculation to overcome the limitation. Top-N news articles, which have high similarity value for a target user, are recommended to the user. To evaluate the proposed news recommendation system, user profiles are generated using Facebook account with participants consent, and we implement a Web crawler to extract news information from PBS, which is non-profit public broadcasting television network in the United States, and construct news profiles. We compare the performance of the proposed method with that of benchmark algorithms. One is a traditional method based on TF-IDF. Another is 6Sub-Vectors method that divides the points to get keywords into six parts. Experimental results demonstrate that the proposed system provide useful news to users by applying user's social network information and WordNet functions, in terms of prediction error of recommended news.

• Journal of Intelligence and Information Systems
• /
• v.26 no.3
• /
• pp.149-169
• /
• 2020

"The Urban Renewal New Deal project", one of the government's major national projects, is about developing underdeveloped areas by investing 50 trillion won in 100 locations on the first year and 500 over the next four years. This project is drawing keen attention from the media and local governments. However, the project model which fails to reflect the original characteristics of the area as it divides project area into five categories: "Our Neighborhood Restoration, Housing Maintenance Support Type, General Neighborhood Type, Central Urban Type, and Economic Base Type," According to keywords for successful urban regeneration in Korea, "resident participation," "regional specialization," "ministerial cooperation" and "public-private cooperation", when local governments propose urban regeneration projects to the government, they can see that it is most important to accurately understand the characteristics of the city and push ahead with the projects in a way that suits the characteristics of the city with the help of local residents and private companies. In addition, considering the gentrification problem, which is one of the side effects of urban regeneration projects, it is important to select and implement urban regeneration types suitable for the characteristics of the area. In order to supplement the limitations of the 'Urban Regeneration New Deal Project' methodology, this study aims to propose a system that recommends urban regeneration types suitable for urban regeneration sites by utilizing various machine learning algorithms, referring to the urban regeneration types of the '2025 Seoul Metropolitan Government Urban Regeneration Strategy Plan' promoted based on regional characteristics. There are four types of urban regeneration in Seoul: "Low-use Low-Level Development, Abandonment, Deteriorated Housing, and Specialization of Historical and Cultural Resources" (Shon and Park, 2017). In order to identify regional characteristics, approximately 100,000 text data were collected for 22 regions where the project was carried out for a total of four types of urban regeneration. Using the collected data, we drew key keywords for each region according to the type of urban regeneration and conducted topic modeling to explore whether there were differences between types. As a result, it was confirmed that a number of topics related to real estate and economy appeared in old residential areas, and in the case of declining and underdeveloped areas, topics reflecting the characteristics of areas where industrial activities were active in the past appeared. In the case of the historical and cultural resource area, since it is an area that contains traces of the past, many keywords related to the government appeared. Therefore, it was possible to confirm political topics and cultural topics resulting from various events. Finally, in the case of low-use and under-developed areas, many topics on real estate and accessibility are emerging, so accessibility is good. It mainly had the characteristics of a region where development is planned or is likely to be developed. Furthermore, a model was implemented that proposes urban regeneration types tailored to regional characteristics for regions other than Seoul. Machine learning technology was used to implement the model, and training data and test data were randomly extracted at an 8:2 ratio and used. In order to compare the performance between various models, the input variables are set in two ways: Count Vector and TF-IDF Vector, and as Classifier, there are 5 types of SVM (Support Vector Machine), Decision Tree, Random Forest, Logistic Regression, and Gradient Boosting. By applying it, performance comparison for a total of 10 models was conducted. The model with the highest performance was the Gradient Boosting method using TF-IDF Vector input data, and the accuracy was 97%. Therefore, the recommendation system proposed in this study is expected to recommend urban regeneration types based on the regional characteristics of new business sites in the process of carrying out urban regeneration projects."

• Journal of Preventive Medicine and Public Health
• /
• v.29 no.4 s.55
• /
• pp.785-799
• /
• 1996

We investigated the 1,004 workers who worked in a automobile factory to study the epidemiologic characterists of dermatoses due to cutting oils. Among the workers, 667(66.4%) answered the questionaire. They are belong to 5 departments of the factory-the Engine-Work(258 workers), Gasoline engine Assembly(210), Diesel engine Assembly(96), Power train Work(86), Power train Assembly(17). We measured the oil mist concentration in air of the departments and examined the workers who had dermatologic symptoms. The results were follows; 1) Oil mist concentration ; Of all measured points(52),9 points(17.2%) exeeded $5mg/m^3$- the time-weighed PEL-and one department had a upper confidence limit(95%) higher than $5mg/m^3$. 2) Dermatologists examined 213 workers. 172 of them complained any skin symptoms at that time - itching(32.5%), papule(21.6%), scale(15.7%), vesicle(12.5%) in order. The abnormal skin site found by dermatologist were palm(29.3%), finger & nail(24.6%), forearm(16.2%), back of hand(8.4%) in order. 3) As the result of physical examination, we found that 160 workers had skin diseases. Contact dermatitis was the most common; 69 workers had contact dermatitis alone(43.1%), 11 had contact dermatitis with acne(6.9%), 10 had contact dermatitis with folliculitis(6.3%), 1 had contact dermatitis with acne & folliculitis, and 1 had contact dermatitis with abnormal pigmentation. Others were folliculitis(9 workers, 5.6%), acne(8, 5.0%), folliculitis & acne (2, 1.2%), keratosis(1, 0.6%), abnormal pigmentation (1, 0.6%), and non-specific hand eczema (47, 29.3%). 4) The prevalence of any skin diseases was 34.0 pet 100 in cutting oil users, and 13.3 per 100 in non- users. Especially, the prevalence of contact dermatitis was 23.0 per 100 in cutting oil users and 23.0 per 100 in non-users. 5) We tried patch test(standard serise, oil serise, organic solvents) on 49 patients to differentiate allergic contact dermatitis from irritant contact dermatitis and found 20 were positive. 6) In a multivariate analysis(independant=age, tenure, kinds of cutting oil), the risk of skin diseases was higher in the water-based cutting oil user and both oil user than non-user or neat oil user(odds ratio were 2.16 and 2.78, respectively). And the risk of contact dermatitis was much higher at the same groups(odds ratio were 5.16 and 6.82, respectively).

• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.1-19
• /
• 2019

Artificial intelligence (AI) is one of the main driving forces leading the Fourth Industrial Revolution. The technologies associated with AI have already shown superior abilities that are equal to or better than people in many fields including image and speech recognition. Particularly, many efforts have been actively given to identify the current technology trends and analyze development directions of it, because AI technologies can be utilized in a wide range of fields including medical, financial, manufacturing, service, and education fields. Major platforms that can develop complex AI algorithms for learning, reasoning, and recognition have been open to the public as open source projects. As a result, technologies and services that utilize them have increased rapidly. It has been confirmed as one of the major reasons for the fast development of AI technologies. Additionally, the spread of the technology is greatly in debt to open source software, developed by major global companies, supporting natural language recognition, speech recognition, and image recognition. Therefore, this study aimed to identify the practical trend of AI technology development by analyzing OSS projects associated with AI, which have been developed by the online collaboration of many parties. This study searched and collected a list of major projects related to AI, which were generated from 2000 to July 2018 on Github. This study confirmed the development trends of major technologies in detail by applying text mining technique targeting topic information, which indicates the characteristics of the collected projects and technical fields. The results of the analysis showed that the number of software development projects by year was less than 100 projects per year until 2013. However, it increased to 229 projects in 2014 and 597 projects in 2015. Particularly, the number of open source projects related to AI increased rapidly in 2016 (2,559 OSS projects). It was confirmed that the number of projects initiated in 2017 was 14,213, which is almost four-folds of the number of total projects generated from 2009 to 2016 (3,555 projects). The number of projects initiated from Jan to Jul 2018 was 8,737. The development trend of AI-related technologies was evaluated by dividing the study period into three phases. The appearance frequency of topics indicate the technology trends of AI-related OSS projects. The results showed that the natural language processing technology has continued to be at the top in all years. It implied that OSS had been developed continuously. Until 2015, Python, C ++, and Java, programming languages, were listed as the top ten frequently appeared topics. However, after 2016, programming languages other than Python disappeared from the top ten topics. Instead of them, platforms supporting the development of AI algorithms, such as TensorFlow and Keras, are showing high appearance frequency. Additionally, reinforcement learning algorithms and convolutional neural networks, which have been used in various fields, were frequently appeared topics. The results of topic network analysis showed that the most important topics of degree centrality were similar to those of appearance frequency. The main difference was that visualization and medical imaging topics were found at the top of the list, although they were not in the top of the list from 2009 to 2012. The results indicated that OSS was developed in the medical field in order to utilize the AI technology. Moreover, although the computer vision was in the top 10 of the appearance frequency list from 2013 to 2015, they were not in the top 10 of the degree centrality. The topics at the top of the degree centrality list were similar to those at the top of the appearance frequency list. It was found that the ranks of the composite neural network and reinforcement learning were changed slightly. The trend of technology development was examined using the appearance frequency of topics and degree centrality. The results showed that machine learning revealed the highest frequency and the highest degree centrality in all years. Moreover, it is noteworthy that, although the deep learning topic showed a low frequency and a low degree centrality between 2009 and 2012, their ranks abruptly increased between 2013 and 2015. It was confirmed that in recent years both technologies had high appearance frequency and degree centrality. TensorFlow first appeared during the phase of 2013-2015, and the appearance frequency and degree centrality of it soared between 2016 and 2018 to be at the top of the lists after deep learning, python. Computer vision and reinforcement learning did not show an abrupt increase or decrease, and they had relatively low appearance frequency and degree centrality compared with the above-mentioned topics. Based on these analysis results, it is possible to identify the fields in which AI technologies are actively developed. The results of this study can be used as a baseline dataset for more empirical analysis on future technology trends that can be converged.