Big data Analytics for Tourism Destination management G. Michael McGrath Professor of Information Systems Victoria University Melbourne

Motivation Travel behavior refer to the actual travel activity of people during their trips such as spatial and temporal movement patterns of tourists. For examples: Where do tourists like to visit? When do tourists visit? What do tourists like or dislike at each of the visited locations? How do tourists travel between places? What routes do they usually take? What activities and events do tourists like to participate in?

Motivation Such knowledge is valuable for: Policy Marker, Government Departments, Business Managers: Destination management Product development Attraction Development and Marketing  Tourism Impact management Transportation Planners: Traffic management Transportation Development.

Motivation Popular methods for capturing travel behavior: Survey and opinion polls Disadvantages: Time consuming Limited in terms of the number of responses Limited in scale of the information captured. Unable to provide comprehensive understanding about the locations, time, interests, movement, etc.

Proposed Technologies Many photo-capturing devices now have built-in global positioning systems (GPS) technology Geotagged photos, with embedded time and geographical information, are shared on social networking websites such as (but not limited to):

Proposed Technologies The geotagged photos have: GPS tag (latitude, longitude) Taken Time Stamp (Date, Month, Year, Hours, Minutes, Second) Textual Metadata (tags, description, title, comments), reflecting what people are interested in. The Actual Photos, provide insight into tourist’s own experience about the entities of interest. People’s Profile (Where they come from?) Allow for comprehensive understanding about tourist behavior without the need of actual engagement.

DEMONSTRATION – Using Flickr Data Large Scale Study of Hong Kong Photo GPS information viewed on Google Earth. (approximately 29,443 photos from 2,100 user) The geotagged photo appear all over the regions of Hong Kong, which indicates that the raw geographical information is noisy and can be misleading if it is directly used for analysis.

DEMONSTRATION Area Of Interest Identifications using Clustering

DEMONSTRATION Movement Trajectory generated from geotagged photos.

Data Driven Approach Data Collection: Data Analysis: Big data sets from Social Network such as: Data Analysis: Develop processing techniques for textual data (review comments), visual data (travel photos), temporal data (travel date and time), location data (GPS coordinates), ect…. Discover Patterns using quantitative data analysis (statistics, data mining)

DEMONSTRATION Tourist traffic flow Analysis Asian Tourists are likely to flow to Hong Kong central area from surrounding areas Tourist from Center Mong Kok tend to visit Tsim Sha Tsui as the next destination while, somes chose to travel directly to Hong Kong Central. Considerable tourist flow are found from Hong Kong Central and Times Square Towers to Tsim Sha Tsui Western tourists tend to visit Hong Kong Central directly right after visiting Center Mong, while, somes travel in the opposite direction. Tourists are more likely to visit Tsim Sha Tsui than Hong Kong Central after visiting Times Square Towers. Some tourists visited Tsim Sha Tsui right after going to the Peak Tower.

DEMONSTRATION Actual Route Taken Analysis: From Center Mong Kok to Time Square Tower Most tourist travel to and from these locations via Tsim Sha Tsui and Hong Kong Central. In detail, they travel along Nathan Road between Centre Mong Kok and Tsim Sha Tsui. Direct ferry services are used to travel between Tsim Sha Tsui and Hong Kong Central. Tourist travel between Hong Kong Central and Time Square Tower on Hennessy Road. Besides, we noticed an interesting finding about tourist preference. There is in fact a direct ferry service from near Times Square Towers to Tsim Sha Tsui. But, most tourist did not chose this line and prefer to travel to Tsim Sha Tsui via the Ferry line from Hong Kong Central.

DEMONSTRATION Time Analysis of Tourist Activity

Demonstration: Location Preference Using Geotagged photos From Flickr Photo Taken by Tourist in Melbourne CBD in July 2015

Demonstration: Location Preference Using Geotagged photos From Flickr Preferred Location to Take sunset photos in Melbourne

Demonstration: Location Preference Using Geotagged photos From Flickr Preferred Location to take Art photos in Melbourne CBD

Demonstration: Outbound Location Preference Top Visited Cities for Australian Travelers.

QUESTIONS 18