• 지능정보연구
• /
• 제23권3호
• /
• pp.95-118
• /
• 2017

최근 도심을 중심으로 연립 다세대의 거래가 활성화되고 직방, 다방등과 같은 플랫폼 서비스가 성장하고 있다. 연립 다세대는 수요 변화에 따른 시장 규모 확대와 함께 정보 비대칭으로 인해 사회적 문제가 발생 되는 등 부동산 정보의 사각지대이다. 또한, 서울특별시 또는 한국감정원에서 사용하는 5개 또는 25개의 권역 구분은 행정구역 내부를 중심으로 설정되었으며, 기존의 부동산 연구에서 사용되어 왔다. 이는 도시계획에 의한 권역구분이기 때문에 부동산 연구를 위한 권역 구분이 아니다. 이에 본 연구에서는 기존 연구를 토대로 향후 주택가 격추정에 있어 서울특별시의 공간구조를 재설정할 필요가 있다고 보았다. 이에 본 연구에서는 연립 다세대 실거래가 데이터를 기초로 하여 헤도닉 모형에 적용하였으며, 이를 K-Means Clustering 알고리즘을 사용해 서울특별시의 공간구조를 다시 군집하였다. 본 연구에서는 2014년 1월부터 2016년 12월까지 3년간 국토교통부의 서울시 연립 다세대 실거래가 데이터와 2016년 공시지가를 활용하였다. 실거래가 데이터에서 본 연구에서는 지하거래 제거, 면적당 가격 표준화 및 5이상 -5이하의 실거래 사례 제거와 같이 데이터 제거를 통한 데이터 전처리 작업을 수행하였다. 데이터전처리 후 고정된 초기값 설정으로 결정된 중심점이 매번 같은 결과로 나오게 K-means Clustering을 수행한 후 군집 별로 헤도닉 모형을 활용한 회귀분석을 하였으며, 코사인 유사도를 계산하여 유사성 분석을 진행하였다. 이에 본 연구의 결과는 모형 적합도가 평균 75% 이상으로, 헤도닉 모형에 사용된 변수는 유의미하였다. 즉, 기존 서울을 행정구역 25개 또는 5개의 권역으로 나뉘어 실거래가지수 등 부동산 가격 관련 통계지표를 작성하던 방식을 속성의 영향력이 유사한 영역을 묶어 16개의 구역으로 나누었다. 따라서 본 연구에서는 K-Means Clustering 알고리즘에 실거래가 데이터로 헤도닉 모형을 활용하여 연립 다세대 실거래가를 기반으로 한 군집분류방법을 도출하였다. 또한, 학문적 실무적 시사점을 제시하였고, 본 연구의 한계점과 향후 연구 방향에 대해 제시하였다.

• 보건교육건강증진학회지
• /
• 제2권1호
• /
• pp.3-50
• /
• 1984

The levels of health knowledge, attitude and practice of housewives considerably effect to the health of households, communities and the nation. This study was designed to grasp the levels of health knowledge, attitude and practice of houswives and analyse the various factors effecting to health in order to provide health education services as well as materials for effective formulation and implementation of health policy to improve the health of the nation. This study has been conducted through interviews by trained surveyers for 4,281 housewives selected from 4,500 households throughout the country for 40 days during July 11-August 20, 1983. The results of survey were analysed by stepwise multiple regression and path analysis are summarized as follows; 1. Based on the measurement instrument applied to this study, the levels of health knowledge, attitude and practice of housewives were extremely low with 54.5 points out of 100 points in full. Higher level with 72 points and above was approximately 21 percent and lower level with 39 points and below was approx. 24 percent. The middle level was approx. 55 percent. In order to implement health programs successively, health education should be more strengthened and to improve the level of health knowledge, attitude and practice (KAP) of the nation, political consideration as a part of spiritual reformation must be concentrated on health. 2. The level of health knowledge indicated the highest points with 57.3 the level of attitude was the second with 55.0 points and the practice level was the lowest with 50.0 point. Therefore, planning and implementation of health education program must be based on the persuasion and motivation that health knowledge turn into practice. 3. Housewives who had higher level of health knowledge, showed their practice level was relatively lower and those who had middle or low level of it practice level was the reverse. 4. Correlations among health knowledge, attitude and practice (KAP) were generally higher and statistically significant at 0.1 percent level. Correlation between total health KAP level and health knowledge was the highest with r=.8092. 5. Health KAP levels showed significant differences according to the age, number of children, marital status, self-assessed health status and concern on health of the housewives interviewed (p<0.001) 6. Health KAP levels also showed significant differences according to the education level, economic status, employment before marriage and grown-up area of the housewives interviewed. (p<0.001) 7. Heath KAP levels showed significant differences according to health insurance benificiary and the existence of patients in the family. (p<0.001). 8. Health KAP levels showed significant differences according to distance to government organizations, schools, distance to health facilities, telephone possession rate, television possession rate, newspaper reading rate and activities of Ban meeting and Women's club. (p<0.001) 9. Health KAP levels showed significant differences according to electric mass communication media such as television, radio and village broadcasting etc. and printed media such as newspaper, magazine and booklets etc., IEC variables such as individual consultation and husband-wife communication, however, there was no significance with group training. 10. Health KAP of the housewives showed close correlation with personal characteristics variables, i.e., education level (r=.5302), age (r=-.3694) grown-up area (r=.3357) and employment before marriage. In general, correlation of health knowledge level was higher than the levels of attitude or practice. In case of health concern and health insurance, correlation of practice level was higher than health knowledge level. 11. Health KAP levels showed higher correlation with community environmental characteristics, Ban meeting and activity of Women's club, however, no correlation with New-village movement. 12. Among IEC variables, husband-wife communication showed the highest correlation with health KAP levels and printed media, electric mas communication media and health consultation in order. Therefore, encouragement of husband-wife communication and development of training program for men should be included in health education program. 13. Mass media such as electric mass com. and printed media were effective for knowledge transmission and husband-wife communication and individual consultation were effective for health practice. Group training was significant for knowledge transmission, however, but not significant for attitude formation or turning to health practice. To improve health KAP levels, health knowledge should be transmitted via mass media and health consultation with health professionals and field health workers should be strengthened. 14. Correlation of health KAP levels showed that knowledge level was generally higher than that of practice and recognized that knowledge was not linked with attitude or practice. 15. The twenty-five variables effecting health KAP levels of housewives had 41 per cent explanation variances among which education level had great contribution (β=.2309) and electric mass com. media (β=.1778), husband-wife communication (β=.1482), printed media, grown-up area, and distance to government organizations in order. Variances explained (R²) of health KAP were 31%, 15%, and 30% respectively. 16. Principal variables contributed to health KAP were education level (β=.12320, β=.1465), electric mass comm. media (β=.1762, β=.1839), printed media, (β=.1383, β=.1420) husband-wife communication (β=.1004, β=.1067), grown-up area and distance to government organizations, in order. Since education level contributes greatly to health KAP of the housewives, health education including curriculum development in primary, middle and high schools must be emphasized and health science must be selected as one of the basic liberal arts subject in universities. 17. Variences explained of IEC variables to health KAP were 19% in total, 14% in knowledge, 9% in attitude, and 10% in health practice. Contributions of IEC variables to health KAP levels were printed media (β=.3882), electric mass comm media (β=.3165), husb-band wife com. (β=.2095,) and consultation on health (β=.0841) in order, however, group training showed negative effect (β=-.0402). National fund must be invested for the development of Health Program through mass media such as TV and radio etc. and for printed materials such as newspaper, magazines, phamplet etc. needed for transmission of health knowledge. 18. Variables contributed to health KAP levels through IEC variables with indirect effects were education level (Ind E=0.0410), health concern (Ind E=.0161), newspaper reading rate (Ind E=.0137), TV possession rate and activity of Ban meeting in order, however, health facility showed negative effect (Ind E=-.0232) and other variables showed direct effect but not indirect effect. 19. Among the variables effecting health KAP level, education level showed the highest in total effect (TE=.2693) then IEC (TE=.1972), grown-up city (TE=.1237), newspaper reading rate (TE=.1020), distance to government organization (TE=.095) in order. 20. Variables indicating indirect effects to health KAP levels were; at knowledge level with R²=30%, education level (Ind E=.0344), newspaper reading rate (Ind E=.0112), TV possession rate (Ind E=.0689), activity of Ban meeting (Ind E=.0079) in order and at attitude level with R²=13%, education level (Ind E=. 0338), activity of Ban meeting (Ind E=.0079), and at practice level with R²=29%. education level (Ind E=.0268), health facility (Ind E=.0830) and concern on health (Ind E=.0105). 21. Total effect to health KAP levels and IEC by variable characteristics, personal characteristics variables indicated larger than community characteristics variables. 22. Multiple Correlation Coefficient (MCC) expressed by the Personal Characteristic Variable was .5049 and explained approximately 25% of variances. MCC expressed by total Community environment variable was .4283 and explained approx. 18% of variances. MCC expressed by IEC Variables was .4380 and explained approx. 19% of variances. The most important variable effected to health KAP levels was personal characteristic and then IEC variable, Community Environment variable in order. When the IEC effected with personal characteristic or community characteristic, the MCC or the variances were relatively higher than effecting alone. Therefore it was identified that the IEC was one of the important intermediate variable.