협력적 필터링 알고리즘을 사용한 개인별 행복감 예측정확도 향상에 대한 연구
A Study on the Improvement of Individual Happiness Predictions Accuracy Using Collaborative Filtering
1
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The purpose of this paper is to improve the prediction accuracy in predicting individual happiness in consideration of individual characteristics and living conditions. Happiness is very subjective, so it should be measured individually. When individual happiness is accurately predicted, a direction for improving happiness can be sought based on that value. Collaborative filtering is an algorithm that builds a recommendation system, and is mainly used to predict the preference in personalized services. Therefore, after separating the respondents into demographic variables, this paper calculated the similarity weight for the satisfaction of living conditions among respondents for each, and applied it to NBCF and CMA, which are collaborative filtering algorithms, to predict the happiness of each respondent, and evaluated the prediction accuracy. As a result, the prediction accuray was higher in the case of using CMA than NBCF, and the case predicted by age group was higher than that predicted by respondent or sex. The more accurately predicted individual happiness obtained as a result of this study is expected to serve as an basic data for deriving the direction of personalized living conditions.
Abstract
The purpose of this paper is to improve the prediction accuracy in predicting individual happiness in consideration of individual characteristics and living conditions. Happiness is very subjective, so it should be measured individually. When individual happiness is accurately predicted, a direction for improving happiness can be sought based on that value. Collaborative filtering is an algorithm that builds a recommendation system, and is mainly used to predict the preference in personalized services. Therefore, after separating the respondents into demographic variables, this paper calculated the similarity weight for the satisfaction of living conditions among respondents for each, and applied it to NBCF and CMA, which are collaborative filtering algorithms, to predict the happiness of each respondent, and evaluated the prediction accuracy. As a result, the prediction accuray was higher in the case of using CMA than NBCF, and the case predicted by age group was higher than that predicted by respondent or sex. The more accurately predicted individual happiness obtained as a result of this study is expected to serve as an basic data for deriving the direction of personalized living conditions.
참고문헌 (12)
- Lee Soojung, "Optimizing Similarity for User-based Collaborative Filtering", Journal of The Korea Society of Computer and Information, 29(11), pp.243~250 (2024)
- 박상준, "행복지표 속에 나타난 행복담론변화연구", 문화경제연구, 25(3), pp.35~60 (2022)
- 변상해, "청소년이 지각하는 주관적 행복감이 공감 능력에 미치는 영향 : 부모-자녀 의사소통의 매개효과를 중심으로", 인문사회 21, 13(5), pp.2973~2988 (2022)
- 박정혜, "직장 여성의 연령대별 주관적 행복감과 관련된 요인 : 청년층과 중년층을 중심으로", Journal of the Korean Data Analysis Society, 24(1), pp.165~178 (2022)
- 이지혜, "자존감 유형에 따른 부정적 사건에 대한 귀인양식 및 부정적 정서의 차이", 한국심리치료학회지, 12(2), pp.53~67 (2020)
- 박새한, "의료 빅데이터로 분석한 만성질환자의 건강정보수준별 차이 연구", 디지털산업정보학회 논문지, 19(4), pp.73~86 (2023)
- 고윤희, "중년 여성의 사회활동 및 신뢰가행복에 미치는 영향 : 삶의 의미 매개효과", Journal of Social Sciences, 63(2), pp.149~172 (2024)
- 오수민, "사회적 변수와 개개인의 감정지수를 함께 고려한 딥러닝 기반 행복 지수 모델 설계", The Journal of the Convergence on Culture Technology, 10(1), pp.489~493 (2024)
- Koo Jeehyun, "The Effect of the Common Evaluation of Customers and Neighbors on the Predictive Error", Journal of the Korean Data Analysis Society, 21(2), pp.745~775 (2019)
- Breese, J. S., "Empirical analysis of predictive algorithms for collaborative filtering", Proc. for the 14th conference on Uncertainty in artificial intelligence, pp.43~52 (1998)
- Lee Heechoon, "Improved algorithm for user based recommender system", Journal of the Korean Data & Information Science, 17(3), pp.717~726 (2006)
- "(2023년 18차 한국복지패널조사)데이터(beta1)"[URL]