摘要
利用数据挖掘的分析方法对用户生成的图片内容(User Generated Content,UGC)进行数据挖掘,已成为旅游市场营销和目的地形象研究的热点。然而,分析和处理海量图片数据需要借助计算机技术,这对于广大旅游研究者而言门槛颇高。本文构建了一个通用的图片数据分析工具,旨在为旅游研究者提供一个海量图片分析平台,利用该数据平台可从4个方面服务于旅游研究:(1)目的地热点(Point of Interest,POI)分析;(2)游客历史对比分析;(3)游客轨迹分析;(4)基于图片的目的地形象感知分析。本文以罗马为例,说明了该分析平台如何应用于旅游研究中,并对利用UGC图片进行旅游研究提供了方法和工具上的支撑。
UGC(User Generated Content),including the social network photos,has been widely adopted in tourism marketing and travel destination image(TDI) research.However,the analysis of large-scale UGC requires some computer technical support,which builds an apparent barrier for ordinary tourism researchers.This paper proposes and implements a general Photo-based Metadata Mining System,named PMMS,aiming to provide a general analysis platform for massive UGC photos.PMMS provides 4 functions to serve the tourism and travel research:(1) Point of Interest(POI) analysis;(2) tourists historical comparison;(3) tourists' movement analysis;and(4) destination image perception based on photos.In POI analysis,the geographic information embedded inside the photos' metadata are extracted and visualized using some Arc GIS tools to achieve the distribution of the points of interest at the destination.In tourists' historical comparison,photos are visualized through their taken time and geographic information,and then a time-line based tourism research could be analyzed using it.Also,as an important topic in tourism research,tourists' movement analysis is frequently adopted in tourism planning and intelligent route design.Lastly,TDI perception is one of the most active topic in tourism research and is the basis of the destination marketing,and UGC photos provide a new dimension for TDI analysis.In this study,we take Rome,the capital of Italy,as a case study to illustrate the capabilities of PMMS in the tourism and travel research.
出处
《旅游导刊》
2017年第6期34-47,共14页
Tourism and Hospitality Prospects
基金
国家自然科学基金青年项目"新媒体语境下旅游形象量化投射模型构建与评估研究"(71502005)
北京市社会科学基金研究基地项目"基于UGC大数据的北京冬奥会海外形象精准投射研究"(17JDGLB006)
北京市教委科技项目"基于图片的北京旅游形象海外精准营销平台建设"(SQKM201710031001)资助
