인턴형 일경험 프로그램 추천 알고리즘에 관한 연구 = A Study on the Recommendation Algorithm for Internship-Type Work Experience Programs

다국어 초록 (Multilingual Abstract)

This study aims to develop and evaluate a recommendation algorithm for intern-type work experience programs. Due to changes in the recruitment environment, demand of youth job seekers for work experience will be increased. This trend will increase as the recruitment environment shifts to experience-based hiring. However, the current work experience management system offers simple search function and does not offer a personalized-recommendation service which help to choose suitable programs.

This study developed a personalized-recommendation system using data from actual participants in an intern-type work experience program. The raw data was integrated and pre-processed in order to analysis and utilization. To compare and evaluate the performance of the developed-recommendation system, core and auxiliary indicators were established. The recommendation algorithm was developed using three different approaches such as item-based collaborative filtering, random forest and a rule-based model using data-driven insights. The three recommended algorithms were compared and evaluated by core and auxiliary indicators, and finally the best algorithm was proposed. This study showed that matching efficiency can be improved by predicting and recommending the preferred programs of participants who wish to participate in an intern-type work experience program.
This study presents guidelines for intern-type work experience recommendation services and suggests for improvements in the current data management system.

목차 (Table of Contents)

• ABSTRACT viii
• 제 1 장 서론 1
• 제 2 장 이론적 고찰 4
• 2.1 청년 일경험 사업 개요 4
• 2.2 정보 필터링 11
• 2.2.1 정보 필터링 개요 11
• 2.2.2 정보 필터링 주요 알고리즘 13
• 2.3 머신러닝 18
• 2.3.1 머신러닝 개요 18
• 2.3.2 머신러닝 주요 알고리즘 20
• 제 3 장 데이터 분석 및 전처리 27
• 3.1 데이터 수집 및 통합 27
• 3.1.1 데이터 수집 27
• 3.1.2 데이터 통합 28
• 3.2 데이터 분석 30
• 3.2.1 참여자 목록 30
• 3.2.2 프로그램 목록 37
• 3.2.3 참여기업 목록 39
• 3.3 데이터 전처리 42
• 3.3.1 결측값 42
• 3.3.2 데이터 변환 43
• 3.4 변수설정 48
• 3.4.1 독립변수 및 종속변수 설정 48
• 3.4.2 변수 유효성 분석 50
• 제 4 장 시스템 개발 57
• 4.1 성능 평가 지표 57
• 4.1.1 핵심 지표 58
• 4.1.2 보조 지표 61
• 4.2 사용언어 및 주요 라이브러리 66
• 4.3 전공 분류 시스템 개발 65
• 4.3.1 전공 분류 알고리즘 68
• 4.3.2 전공 분류 시스템 개발 69
• 4.3.3 전공 분류 시스템 개발 결과 71
• 4.4 추천 시스템 개발 72
• 4.4.1 협업 필터링 72
• 4.4.2 랜덤포레스트 78
• 4.4.3 규칙기반 83
• 4.5 성능 평가 90
• 제 5 장 결론 96
• 참고문헌 99
• 부 록 104