Evaluating the Travel Distances of Events Visitors and Regular Visitors Using Mobile Positioning Data: The Case of Estonia

Abstract

Event tourism has become an important economic component of modern tourism, given the growing numbers of visitors and the development of local communities that it entails. This study examines whether the distances traveled by foreign visitors to events differ from those traveled by non-event (i.e., regular) visitors, and which factors influence such distances. We use passive mobile positioning data obtained from visitors to 145 events in Estonia (119,288 visits) and from a comparison group of regular visitors (199,859 visits) between 2006 and 2009. The results show that events visitors come from nearer locations than regular visitors; similarly to regular visitors, the distance decay principle applies in case of events visitors; and that events bring more visitors from more distant countries in the off-season (winter).

Measuring tourism destinations using mobile tracking data

“…For example, Saaremaa functions as a destination for Finns and Swedes while Tartu County attracts more Latvians. The overall visitation shares are in accordance with the principle of distance decay in Estonia (Nilbe et al, 2014). Analysis of the top seven nationalities also revealed temporal preferences: German visitors are very summer-oriented, whereas Russians are relatively insensitive to season in all three destination areas.…”

“…When analysing the country of origin, one of the most important factors influencing the number and composition of tourists is the distance decay principle, in that more visitors come from nearby countries (McKecher & Lew, 2003;Olsson, 1965). Accordingly, more visitors come to Estonia from countries that are geographically closer to it, such as Finland and Latvia (Nilbe et al, 2014).…”

“…The use of tracking technologies in travel and tourism studies has seen rapid development, beginning in the early 2000s with explorative and conceptual studies (e.g., Ahas & Mark, 2005;Asakura & Hato, 2004), followed by numerous descriptive studies exploiting new data sources (e.g., Ahas, Aasa, Mark, Pae, & Kull, 2007;Shoval, 2008) and the use of tracking data to analyse more sophisticated aspects of tourists' spatial behaviour (e.g., McKercher, Shoval, Ng, & Birenboim, 2012;McKercher, Shoval, Park, & Kahani, 2015;Nilbe et al, 2014;Shoval et al, 2011). Compared with traditional accommodation and survey data, GPS and mobile phone based tracking technologies enable us to study tourism more precisely and effectively because (a) the spatial and temporal accuracy of the data are better; (b) the tracking periods are longer; (c) tracking allows us to follow a tourist throughout his/her visit; and (d) digital data collection and processing are easy and timeliness.…”

“…One possibility for measuring visitor flows in destinations quantitatively is via the use of different ICT-based tracking technologies. Various tracking datasets have been applied to the study of tourism and tourist movements, including mobile positioning data (e.g., Ahas, Aasa, Roose, Mark, & Silm, 2008;Calabrese & Ratti, 2006;Nilbe, Ahas, & Silm, 2014), GPS data (e.g., Grinberger, Shoval, & McKercher, 2014;Shoval & Isaacson, 2007;Shoval, McKercher, Ng, & Birenboim, 2011), bluetooth data (e.g., Versichele et al, 2014;Yoshimura et al, 2014), user-generated data such as geo-located tweets from Twitter (e.g., Hawelka et al, 2014), and georeferenced photos from the photo-sharing webpage Flickr (e.g., Girardin, Fiore, Ratti, & Blat, 2008).…”

Extracting regular mobility patterns from sparse CDR data without a priori assumptions

“…However, we think that this should only be done if the CDR are temporally dense enough. Other work on CDR has focussed on extracting derived information, such as home and work locations (Ahas et al 2010;Becker et al 2013;Eagle, De Montjoye, and Bettencourt 2009;Isaacman et al 2011) or classifications of users (Furletti et al 2013;Nilbe, Ahas, and Slim 2014). In their studies, Ranjan et al 2012) have found that using CDR as a sparsely and inaccurately sampled trajectory incurs biases, especially for users that do not have many CDR distributed evenly through the day.…”