Evaluation of the Forecast Models of Chinese Tourists to Thailand Based on Search Engine Attention: A Case Study of Baidu

摘要

Tourism, as a rather complex behaviour, includes lots of stages during the course of decision-making. Currently, related people have considered a lot of models for the tourism demand prediction. The research described in this paper aims at using the Baidu trends based on internet big data to construct inflow index of Chinese tourists to Thailand to provide forecasts on auxiliary. By limiting keywords set to only travel related keywords, dealing with keywords with different weights according to relations to series of interests, Baidu variable is constructed. We compare a number of standard models with Baidu-augmented models, and then evaluate if the variable of Baidu has raised these models’ prediction performances. We tested for the seasonal unit roots and the result confirmed that there were no seasonal unit roots. The evaluation result show models including Baidu variables can improve forecasting accuracy significantly and the choice of exogenous variables may critically affect prediction ability.