Discovering Landmark Preferences and Movement Patterns from Photo Postings

Abstract

This article presents a geovisual analytics approach to discovering people's preferences for landmarks and movement patterns from photos posted on the Flickr website. The approach combines an exploratory spatio-temporal analysis of geographic coordinates and dates representing locations and time of taking photos with basic thematic information available through the Google Maps Web mapping service, and interpretation of the analyzed area. The article describes data aggregation and filtering techniques to reduce the size of the dataset and focuses on information addressing research questions. The results of analysis for the Seattle metropolitan area help to distinguish between sites that are occasionally popular among the photographers and can be considered as potential attractions from sites that are regularly visited and already known as city landmarks. The analysis of photographers' movements across the metropolitan area shows that most photographers' itineraries are short and highly localized.

Opportunities for Volunteered Geographic Information Use in Spatial Planning

“…In the representation model, SMGI may be used to facilitate the description of a geographic context, providing experiential knowledge that is usually dismissed in official information and integrating A-GI with a pluralist vision of geographic phenomena, which may be used to identify social and cultural dynamics affecting the area. For example, SMGI from several Location-Based Social Networks (LBSNs) has been used to identify the most appreciated Points of Interest (POIs) and landmarks in a study area (Jankowski et al, 2010), the pedestrian paths in the historical centre of a city, the neighbourhoods featuring the lowest number of services and the different land uses in an urban environment (Frias-Martinez et al, 2012), and to classify urban areas (Noulas et al, 2011).…”

“…For instance, SMGI may be extracted for studying the movements of users in urban environments (Jankowski et al, 2010), the utilisation rates of public spaces (Torres and Costa, 2014) and the neighbourhood perceptions of users (Massa and Campagna, 2014), as well as the dynamics of different population groups (Longley et al, 2015).…”

Towards (Re)Constructing Narratives from Georeferenced Photographs through Visual Analytics

“…For example, Zeng et al (2012) offer a method to discover 'rational paths' for tourists, Jankowski et al (2010) use a geovisual analytics approach to explore 'landmark preferences' based on movement patterns and Andrienko et al (2010) propose a set of visual analytics methods to discover and reconstruct 'place histories' from people's activity traces obtained from photo collections.…”

Gaining Knowledge from Georeferenced Social Media Data with Visual Analytics

“…by a regular (Andrienko et al 2009) or irregular (Jankowski et al 2010;Andrienko et al 2012) grid, and the photo taking events are aggregated by these compartments and time intervals. The resulting time series of the event counts are visualised on a map (Andrienko et al 2009) or on a time graph (Jankowski et al 2010;Andrienko et al 2012), which is linked to a map display through interactive techniques, including synchronous highlighting, selection, and filtering of corresponding visual objects. By analysing the time series using either mostly interactive (Jankowski et al 2010) or computationally supported (Andrienko et al 2012) techniques, the researchers detected places with interesting temporal patterns of visits, such as periodic peaks at particular times of the year, very high irregularly occurring peaks, and significant increase of place popularity starting from a particular time.…”

“…To study the spatial distribution of the photos over a territory and compare the temporal patterns of visiting different parts of it, the territory is divided into compartments, e.g. by a regular (Andrienko et al 2009) or irregular (Jankowski et al 2010;Andrienko et al 2012) grid, and the photo taking events are aggregated by these compartments and time intervals. The resulting time series of the event counts are visualised on a map (Andrienko et al 2009) or on a time graph (Jankowski et al 2010;Andrienko et al 2012), which is linked to a map display through interactive techniques, including synchronous highlighting, selection, and filtering of corresponding visual objects.…”