ENTER 2015 Research Track Slide Number 1

Linked Data for Cross-Domain Decision-making in Tourism

Marta Sabou, Adrian M.P. Brașoveanu, & Irem Önder

MODUL University Vienna, Austria {marta.sabou, adrian.brasoveanu,irem.onder} @modul.ac.at

http://www.modul.ac.at

ENTER 2015 Research Track Slide Number 2

Agenda

• Motivation • Purpose of the study• Creating the ETIHQ linked data dataset

and dashboard• Summary • Lessons learned

ENTER 2015 Research Track Slide Number 3

Motivation• Tourism decision making (i.e. benchmarking,

forecasting) does not only depend on pure tourism statistics such as arrivals, bednights and capacity) but must also consider data from other domains.

• For example:– Economic indicators (i.e. Inflation rate, unemployment

rate, currency exchange rate) are interlinked with tourism consumption.

– Environmental indicators (i.e. CO2 emissions) could be affected in heavily touristic areas.

ENTER 2015 Research Track Slide Number 4

Motivation cont.• Most tourism decision support systems

usually only cater for investigating tourism indicators in isolation from economic or sustainability indicators.

• Semantic Web and Linked Data (LD) technologies have been developed to support the intelligent integration of data on the Web (Berners-Lee et al., 2001).

ENTER 2015 Research Track Slide Number 5

What is Linked Data?

• Linked Data technologies provide a mechanism to publish “intelligent” data on the Web and creating links between elements of different datasets, thus facilitating their integration.

ENTER 2015 Research Track Slide Number 6

Purpose of the study

• To create the technical solution for providing TourMIS (tourism related database) as LD

• To implement a cross-domain decision support dashboard

• To reflect on the main lessons learned while using LD technology for tourism.

ENTER 2015 Research Track Slide Number 7

Related Work

• Sources of Tourism Indicators: UN’s World Tourism Organization (UNWTO), Eurostat, TourMIS, The World Bank.

• Tourism Decision Support Systems: TourMIS, PATA, Bastis

ENTER 2015 Research Track Slide Number 8

Sources of Tourism indicators

UNWTO Eurostat WorldBank ETIHQ

Country Y Y Y Y

City N N N Y

Year Y Y Y Y

Month N Y N Y

ENTER 2015 Research Track Slide Number 9

Tourism decision support systems

Coverage Indicator’s domain

Technology

Bastis Baltic SeaRegion

Mixed Wiki/community contributions

Pata Asia Tourism Not known (commercial)

ETIHQ Europe Tourism, economics,environment

LD (QB)

ENTER 2015 Research Track Slide Number 10

Creating the ETIHQ Linked Data Dataset

• Data Cleaning• Semantic Modeling• RDB2RDF conversion• Interlinking• Linked Data Interface Publishing

ENTER 2015 Research Track Slide Number 11

TourMIS - a key source of European Tourism Statistics

Developed at

… supported and used by several national and European tourism

organisations

Data since 1985

10+ indicators

About 154 destinations

Annual and monthly measurements

Heterogeneous ownership

Daily Updates

ENTER 2015 Research Track Slide Number 14

Indicators in ETIHQ• Economic Indicators

– GDP growth (annual %);– Inflation, consumer prices (annual %);– Consumer price index (2005 = 100);– Official exchange rate (LCU per US$, period avg);– Unemployment rate, total (%of total labor force).

• Sustainability Indicators– CO2 emissions (kt);– Forest area (% of land area);– Roads, paved (% of total roads); – Agricultural land (% of land area).

ENTER 2015 Research Track Slide Number 15

Example: German tourists in Venice, Budapest and Dubrovnik

ENTER 2015 Research Track Slide Number 16

Adding a new indicator to the list

ENTER 2015 Research Track Slide Number 17

Adding GDP to the chart

ENTER 2015 Research Track Slide Number 18

Summary

• In this paper we described advances to the state of the art in terms of

(1) publishing TourMIS data as linked data;

(2) interlinking TourMIS data with data form other data sources covering the economic and statistical domains;

(3) creating a visual dashboard that explores this integrated data to support cross-domain decision making processes.

ENTER 2015 Research Track Slide Number 19

Lessons Learned

• LD technologies, greatly facilitate data integration at the syntactic and semantic level to establish links between various datasets.

• Data compatibility (had to make changes to the TourMIS dataset before publishing it).

• Licensing problems due to the heterogeneous origin of the TourMIS data set.

ENTER 2015 Research Track Slide Number 20

Thank you!irem.onder@modul.ac.at

marta.sabou@modul.ac.at

adrian.brasoveanu @modul.ac.at