REVEALING TOURIST HOTSPOTS IN YOGYAKARTA CITY BASED ON SOCIAL MEDIA DATA CLUSTERING

GeoJournal of Tourism and Geosites ISSN 2065-1198, E-ISSN 2065-0817 Year XIV, vol. 34, no. 1, 2021, p.218-225 DOI 10.30892/gtg.34129-640 REVEALING TOURIST HOTSPOTS IN YOGYAKARTA CITY BASED ON SOCIAL MEDIA DATA CLUSTERING Totok Wahyu WIBOWO Faculty of Geography, Universitas Gadjah Mada, Department of Geographic Information Science, Yogyakarta, Indonesia, e-mail: Sigit Heru Murti Budi SANTOSA* Universitas Gadjah Mada, Faculty of Geography, Department of Geographic Information Science, Yogyakarta, Indonesia, e-mail: Bowo SUSILO Universitas Gadjah Mada, Faculty of Geography, Department of Geographic Information Science, Yogyakarta, Indonesia, e-mail: Taufik Hery PURWANTO Universitas Gadjah Mada, Faculty of Geography, Department of Geographic Information Science, Yogyakarta, Indonesia, e-mail: Citation: Wibowo, T.W., Santosa, S.H.M.B., Susilo, B., & Purwanto, T.H. (2021). REVEALING TOURIST HOTSPOTS IN YOGYAKARTA CITY BASED ON SOCIAL MEDIA DATA CLUSTERING. GeoJournal of Tourism and Geosites, 34(1), 218–225. https://doi.org/10.30892/gtg.34129-640 Abstract: Cities have a common characteristic in the form of land utilisation, which is dominated by built -up areas. Tourism is an essential aspect of city development because it can involve the identity of the city. Historical buildings, landmarks, shopping centres and museums are generally interesting places for tourists to visit. Yogyakarta, the research area, is synonymous as a city of culture and of students. Knowledge of the spatial clustering patterns of tourists can be one of the references for urban development. Social media data were used in the study as an alternative to direct data collection, which requires considerable resources. Flickr and Twitter were used as proxies to dete rmine the distribution of tourists, and the DBSCAN and HDBSCAN clustering algorithms were used to determine the centres of tourist activity. Further more, Flickr data were analysed temporally to determine the impact of the COVID-19 pandemic on tourism in Yogyakarta City. The clustering of social media data results shows that there are several city hotspots, besides the already well-known tourist attractions. Apart from city landmarks, several other tourist hotspots were revealed through the clustering process, such as accommodation, shopping centres, entertainment venues and souvenir shops, which also support tourism activities. The impact of COVID-19 on tourism in Yogyakarta City can be reflected through the number of uploaded photos by tourists on Flickr, which has decreased since March 2020. Key words: Social Media, Flickr, Twitter, HDBSCAN, COVID-19 * * * * * * INTRODUCTION Tourism in urban areas has certain characteristics because most of the land use is generally built -up areas. Various human-made landmarks, such as buildings, shopping centres, recreation zones, or iconic city locations can be beautiful places for tourists to visit (Hu et al., 2015). This condition is different from tourist attractions clearly defined as having a particular space and function. Because eac h city is unique, tourists can visit exciting places there. The problem is that in public areas, it is likely that tourism officers will not record the number of visitors. Several previous studies have called this tourist perception the ‘vague place’ concept (Montello et al., 2014; Montello et al., 2003). Urban tourism can generate much diversification of tourist choices, so it is not easy to decide where to visit when on short visits (Jansen-Verbeke, 1986; Salas-Olmedo et al., 2018). Apart from visits to exciting places, tourists also need adequate supporting infrastructure. Some of this may also become tourist destinations in its own right, because it offers absolute uni queness. For example, in terms of hotel selection, tourists tend to choose hotels close to their destination, generally within walking distance (Shoval et al., 2011). This makes a tourist destination attractive because determination of a tourist centre is not only based on officers' counts at officially managed tourist sites. Knowledge of tourists' favourite locations is required for more effective city management, espe cially if there are new locations that offer opportunities for further exploration (Devkota et al., 2015). The number of incoming tourists is a good indicator to assess the competitiveness of regional tourism. There are three sources of incoming tourist number data, namely: statistical records, travel records, and social media (Mou et al., 2019). Statistical records can provide accurate incoming tourist data but are unable to show the spatial distribution of tourists. Meanwhile, the use of travel notes has two main challenges, namely the existence of incomplete data and an inaccurate report by the author. Social media provide opportunities for passive tourist data acquisition due to its increasing popularity (Girardin et al., 2008; Mou et al., 2019; Önder et al., 2014). The digital footprint left by tourists through social media content opens up opportunities for spatial tourist studies. In general, studies related to tourist concentration can be conducted using census techniques or surveys based on samples (Salas-Olmedo et al., 2018). However, both methods require considerable resources and do not necessarily cover a broad enough spatial or temporal scope. Big data offers new tourism res earch opportunities by providing high levels of spatial and temporal data to analyse large volumes of tourist spatio-temporal patterns (Goodchild, 2007). Big data for tourism studies can be obtained from three sources: users, devices and operations (Li et al., 2018). These three big data sources for tourism are influenced by developments in social media platforms, the Internet of Things (IoT), and services for tourism-related operations (e.g., web searches, web page visits and online ticket purchases). Based on big data from these three primary sour ces, tourist and tourism market behaviour can be better explored and understood by academia and industry. For example, big data on a large scale make it possible to overcome the limitations of survey data user sample size problems and provide new ways to understand tourist beha viour (Yang et al., 2015). Big data analysis is also known to provide sufficient data without sample bias to understand such behaviour (Li et al., 2017). The use of geotagged photos from photo-sharing services or other social media can enrich tourism data sources and can be used in tourism * Corresponding author http://gtg.webhost.uoradea.ro/ Revealing Tourist Hotspots in Yogyakarta City Based on Social Media Data Clustering planning and management. The density of geotagged photos shows the distribution of the tourist presence across a city, and is easier to perform than direct measurements (García-Palomares et al., 2015). The direct benefit of density mapping is that it can be used to determine spatiotemporal clusters (Hu et al., 2015; Kisilevich et al., 2013) and id (...truncated)