모바일폰 위치기반 생활이동 빅데이터를 활용한 통행목적별 도시활력 영향요인 분석 : PageRank 알고리즘과 SHAP 기계학습을 활용하여Analysis of Determining Factors of Urban Vitality with Mobile Phone Location-Based Origin-Destination Bigdata by Travel Purpose : Using the PageRank Algorithm and SHAP Machine Learning
- Other Titles
- Analysis of Determining Factors of Urban Vitality with Mobile Phone Location-Based Origin-Destination Bigdata by Travel Purpose : Using the PageRank Algorithm and SHAP Machine Learning
- Authors
- 박준상; 김선재; 이수기
- Issue Date
- Oct-2022
- Publisher
- 대한국토·도시계획학회
- Keywords
- Urban Vitality; Mobility of Living Population; PageRank; Multi-layer Perceptron; Interpretable Machine Learning; 도시활력; 생활인구 이동; PageRank; 다층 퍼셉트론; 해석가능한 기계학습
- Citation
- 국토계획, v.57, no.5, pp 72 - 89
- Pages
- 18
- Indexed
- KCI
- Journal Title
- 국토계획
- Volume
- 57
- Number
- 5
- Start Page
- 72
- End Page
- 89
- URI
- https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/185592
- DOI
- 10.17208/jkpa.2022.10.57.5.72
- ISSN
- 1226-7147
2383-9171
- Abstract
- 본 연구는 서울시 생활이동 데이터와 PageRank 알고리즘을 활용하여 도시활력이라는 개념에 대해 통행의목적에 따라 통근 통행과 비통근 통행으로 구분하여 조작적으로정의하였다. 이는 통근 통행과 비통근 통행의 성격이 서로 다르기 때문이다(민병학 외, 2016). 통행의 목적별로 통근 통행은 출퇴근, 등하교 등과 같이 뚜렷한 목적을 가지고 있는 통행이며, 비(非)통근 통행은 통근 통행을 제외한 대부분의 통행을 포함한다.
이러한 비통근 통행은 일상생활 활동이 일어나는 근린 지역과 밀접한 연관을 가지며, 근린 환경 변수에 영향을 많이 받는다(박영준·박소현, 2019; Krizek, 2003; 이남휘·최창규, 2020). 또한, 도시활력에 영향을 미치는 요인을 분석하기 위해 주요 토지이용을 대변할 수 있는 POI 관심시설, 건축물, 교통환경, 가로경관 등을 조작화하여 분석에 활용하였다.
분석 방법론으로는 각 변수가 도시활력과 가지는 선형 또는 비선형 관계를 도출하기 위해 해석가능한 기계학습을 활용하였다.
나아가 분석 결과를 통해 통행목적별 도시활력을 증진시키기 위한 정책적 시사점을 도출하였다.
Urban vitality is an important indicator for evaluating a city’s sustainability. Urban vitality increases when a social space emerges where people can interact with each other in a city. Although many studies have tried to measure urban vitality and its determining factors, few studies have measured it using moile phone location-based, origin-destination (OD) big data. The aim of this study is to analyze the determining factors of urban vitality with mobile phone big data using the PageRank algorithm and interpretable machine learning techniques. The focus is the nonlinear relationships between urban vitality and its determining factors. The main results of the analysis are as follows. First, urban vitality according to moile phone location-based, OD big data by travel purpose has different determining factors. For instance, while the perception of street scenery had a considerable influence on the urban vitality of non-commuting travel, it had no impact on the urban vitality of commuting travel. Second, restaurant Point of Interest (POI) density and subway station exit density had positive associations with urban vitality for both leisure and utility travel purposes. Third, street safety was a significant variable for urban vitality, regardless of travel purposes of the population. This finding indicates that the safety of the street environment encourages urban vitality. Finally, the interpretable machine learning analysis indicated that the relationships between urban vitality and its determining factors were nonlinear. Overall, the study findings demonstrate the useful application of mobile phone, location-based, OD big data to examine urban vitality and provide specific policy implications for promoting it.
- Files in This Item
-
Go to Link
- Appears in
Collections - 서울 공과대학 > 서울 도시공학과 > 1. Journal Articles
![qrcode](https://api.qrserver.com/v1/create-qr-code/?size=55x55&data=https://scholarworks.bwise.kr/hanyang/handle/2021.sw.hanyang/185592)
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.