Upload
ifitt
View
197
Download
0
Embed Size (px)
Citation preview
ENTER 2014 Research Track Slide Number 1
Constructing a Data Warehouse Based Decision Support Platform for
China Tourism IndustryXiangjie Qiao, Lingyun Zhang, Nao Li and Wei Zhu
Institute of TourismBeijing Union University, China
ENTER 2014 Research Track Slide Number 2
Content• Background
• Related research
• System introduction
• Conclusion and future work
ENTER 2014 Research Track Slide Number 3
Background• in-depth studies on changes and impacts of big data from the view
of management and intelligent decision-making are very limited (Feng, Guo, Zeng, Chen & Chen, 2013)
• The application of big data in the tourism industry, particularly by
tourism enterprises, is even more limited and very rare in
industrial development policy making by government– FlightCaster analyzed the past ten years of flight data to forecast whether a flight
would be late (Mayer & Kenneth,2013)
– Teradata eCircle company used data and applications to keep London moving
during the 2012 Olympic and Paralympic Games (McDonald, 2013)
ENTER 2014 Research Track Slide Number 4
Background• China’s tourism public management and
service is facing a big challenge– The numbers of tourists often exceeds the capacity
of popular tourist attractions– Individual backpack travelers often face an
information flood; they needs more personalized information and services
– Some problems like resource depletion, pollution, and worsening ecological environment have become the key issues for industry’s sustainable development
ENTER 2014 Research Track Slide Number 5
Background• Data is rich for decades of industry development
(data rich but information poor)• BI has been effectively used in enterprises for
managers’ decision making, which can also applied in government policy making to assist tourism public management and service
• Make full use of the data exists in tourism industry– Tourism safety emergency command– Tourism demand forecasting– Tourism resources carrying capacity monitoring– Public information services
ENTER 2014 Research Track Slide Number 6
Content• Background
• Related research
• System introduction
• Conclusion and future work
ENTER 2014 Research Track Slide Number 7
Related research• Data Warehouse for tourism industry
– Although huge amounts of data are available at tourism
destinations, these valuable knowledge sources typically remain
unused (Danubianu, Socaciu & Barila, 2009; Hendawi & El-Shishiny, 2008)
– There exists high demand in electronic publishing of market research
results in the tourism industry (Wӧber, 1998)
– Some tourism data warehouse prototypes were built with various
indicators. (Hendawi & El-Shishiny,2008; Hӧpken, Fuchs, Hӧll, Keil & Lexhagen, 2013)
– It is still not popular to develop decision support systems for government
or industry managers
ENTER 2014 Research Track Slide Number 8
Related research• Data mining application in tourism industry
– automatic predicting trends or behaviors , association analysis , clustering, concept description and deviation detection
• Tourism demand forecast– traditional methods: delphi, time series, econometric
methods– modern methods: artificial neural network, rough set, fuzzy
time series, grey theory and support vector machine; Neural network often outperform others
– compared to the traditional econometric or statistical modeling techniques, data mining is still at its infancy(Law et al., 2007)
• Other methods are more widely used for tourism marketing: market segment , cross selling, and CRM
ENTER 2014 Research Track Slide Number 9
Related research
• Data mining application in tourism industry– Web mining is to find interesting patterns from
hyperlinks, web content and web logs (Wӧber, 2007; Pitman, Zanker, Fuchs & Lexhagen, 2010)
• building a personalized web site• clustering customers• improving customer loyalty• destination or products recommendations
ENTER 2014 Research Track Slide Number 10
Content• Background
• Related research
• System introduction
• Conclusion and future work
ENTER 2014 Research Track Slide Number 11
Technical architecture
ENTER 2014 Research Track Slide Number 12
Indicators and data sourcePrimary Indicators
Secondary Indicators Data Source
market inbound tourism market national tourism administrationoutbound tourism market national tourism administrationdomestic tourism market national tourism administrationinternational tourism market UNWTO, tourism official web site of
the country
industry employment enterprise direct reporting systemstourism investment national tourism administration,
enterprise direct reporting systemsaccommodation for visitors in hotels and similar establishments
enterprise direct reporting systems
travel agencies enterprise direct reporting systems
tourist attractions enterprise direct reporting systems
listed companies listed companies' Quarterly report
economy GDP, exchange rate, per capita disposable income, international economic indicators
national bureau of statistics, the world bank, IMF, the people's bank of china, OECD, UNCTAD
ENTER 2014 Research Track Slide Number 13
Implementation process
ENTER 2014 Research Track Slide Number 14
Implementation process
The logic structure of arrivals for accommodation
ENTER 2014 Research Track Slide Number 15
Implementation process
Dimension level for regions/countries
ENTER 2014 Research Track Slide Number 16
Application cases
Multi-dimensional report viewed by Five Brics
ENTER 2014 Research Track Slide Number 17
Application cases
Example for the correlation between two charts
ENTER 2014 Research Track Slide Number 18
Application cases
Drill path analysis in management cockpit
ENTER 2014 Research Track Slide Number 19
Content• Background
• Related research
• System introduction
• Conclusion and future work
ENTER 2014 Research Track Slide Number 20
Conclusion and future work
• The greatest difficulty doesn’t lie at the technical level, but rather at the needs analysis level to organize the data well
• Data preparation and validation are tedious tasks during the platform’s implementation
• The system is still in development, evaluations should be done to test its effectiveness and get some revisions
ENTER 2014 Research Track Slide Number 21
Conclusion and future work• To make full use of the data
– Developing applications such as tourism early warning and monitoring system, tourism demand forecasting
– Modeling some indexes of the industry such as industry climate index, tourism competitiveness index, tourism purchasing index
• To collect more external data into the data warehouse
ENTER 2014 Research Track Slide Number 22
Xiangjie Qiao
Institute of TourismBeijing Union University, China