Practice of Taipei Tour Planning System Based on Attribute Classification and Time Saver Algorithm

As the Metro network has better coverage and connectivity, Taipei city which has been known for its wide variety of cuisines and snacks now offers much convenience for tourists and backpackers. Most tour information providing or scheduling software in the market provides basic query or search functions, but their output is fixed and lack of customized recommendation based on tourists´ personal characteristics. If travelers have no prior knowledge of their target destination, they may be flooded by information from guidebooks or the Internet and have trouble in making arrangements. Therefore the study makes a practice on tour scheduling recommendation based on two popular classification models -- Logistic Regression and KNN. Meanwhile, a time saver algorithm is proposed to offer an efficient and reasonable route for travelers with suggested spots. The classification models offer recommendations with accuracy rate of 74.5% and 63.5%, respectively. And routes organized by time savor algorithm outperform that by other algorithms in terms of rationality from the opinions of our questionnaire respondents. This prototype system is implemented by integration of MS Excel spreadsheet, Visual Basic for Applications and R-Excel to offer better usability and expandability with the cost of only Excel itself.