Recommendation systems in the context of tourism
Kämppä, Joni (2016)
Kämppä, Joni
Haaga-Helia ammattikorkeakoulu
2016
All rights reserved
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2016121320249
https://urn.fi/URN:NBN:fi:amk-2016121320249
Tiivistelmä
Tourism industry has grown despite recent different global issues, like economic crisis. The industry is steadily growing each year globally. Lot of travelling related information is avail-able for the travelers in form of blogs, travelling sites and applications. The search for potentially interesting sights to visit, during the trip, might become tedious task. The information sources are extensive, and the number of different establishments and tourist attractions are noteworthy. Whilst travelling abroad, the tourists don’t have good information available, data roaming costs make the use of smartphones avoidable, and access to reliable and fast wireless internet-connections might vary between countries and cities. There are mobile applications available to be used as tour guides, with offline-access to location specific information, but these usually require one application per city or country. These applications are also lacking the intuition with their recommendations, and knowledge of their user, and can offer only static, predefined content.
This thesis focuses on the research for the benefits of using machine learning techniques and recommendation systems to simplify the trip planning process for the end user. Various methods could be utilized to find patterns in given user’s past behavior or to find similarities and correlations between known points of interest. These smart systems could be used to give personally tailored experience and recommendations for user. These recommendations could be dependent on their schedule, preferences and budget. Machine learning would also automate the travel planning process and make the finding of the new interesting places easier for wider audiences.
The research contains background study about the tourism industry, technological aspects and describes the general idea behind machine learning. Background study is followed by closer look at the different machine learning techniques and discusses their potentiality to be used in the given context. Suitability of different machine learning algorithms is analyzed in the empirical stage of the process
The last parts of this thesis describe the process of gathering sample data. Sample data is analyzed to find the best ways to use it as training data. Three different machine learning models were constructed to find out how well the test data could be classified. Predictions would be done in context of user’s preferences for places worth of visit.
This thesis focuses on the research for the benefits of using machine learning techniques and recommendation systems to simplify the trip planning process for the end user. Various methods could be utilized to find patterns in given user’s past behavior or to find similarities and correlations between known points of interest. These smart systems could be used to give personally tailored experience and recommendations for user. These recommendations could be dependent on their schedule, preferences and budget. Machine learning would also automate the travel planning process and make the finding of the new interesting places easier for wider audiences.
The research contains background study about the tourism industry, technological aspects and describes the general idea behind machine learning. Background study is followed by closer look at the different machine learning techniques and discusses their potentiality to be used in the given context. Suitability of different machine learning algorithms is analyzed in the empirical stage of the process
The last parts of this thesis describe the process of gathering sample data. Sample data is analyzed to find the best ways to use it as training data. Three different machine learning models were constructed to find out how well the test data could be classified. Predictions would be done in context of user’s preferences for places worth of visit.