Smart Travel Recommender
In 2017, the Asia-Pacific region welcomed 324 million tourists, close to a quarter of the world’s total. Growth in this region is expected to continue unabated over the coming years at a strong 6% annually, despite the current halt in travel due to the Coronavirus Pandemic. Due to the undisputed popularity of travel, there has been a boom in AI-powered travel planners like Anywhr and TripHobo which can recommend destinations and plan itineraries. However, such planners tend to focus on a particular userbase, making them a somewhat niche product. In order to design a more inclusive and dynamic approach to trip planning, our team proposes a novel travel recommendation system that uses the internet search history of users to form a profile of their personality and preferences. This data can then be manipulated to make highly accurate and personalized travel itineraries.
We have scraped data across hundreds of webpages to create a model that can recognize what kinds of websites you visit by their content. To build a sizable database of attractions to choose from, we have also collected a mountain of holiday attraction data from travel sites such as TripAdvisor. With these, we have developed a system that is capable of understanding the kind of holiday you will be interested in, simply by accessing your web browsing history.
Full Name | Student ID | Areas of Responsibility |
---|---|---|
Yee Zhi Quan Darrel | A0213571M | Web-scraping Algorithms, NLP Pre-processing |
Yang Jieshen | A0003901Y | User Profiling Model |
Onn Wei Cheng | A0092201X | UI Development, Rule-based Systems |
Cheng Kok Cheong | A0038791W | Attraction Search Algorithms |
For a full copy of the user guide, please refer to Appendix III in the group report.
Please refer to ProjectReport folder in the repository.