期刊名称：International Journal of Statistics and Applications
印刷版ISSN：2168-5193
电子版ISSN：2168-5215
出版年度：2021
卷号：11
期号：3
页码：61-69
DOI：10.5923/j.statistics.20211103.02
语种：English
出版社：Scientific & Academic Publishing Co.
摘要：We investigated how grouping consumers with similar interests is important for revenue optimization. A real dataset application is carried out to see this importance. To identify traveler archetypes from Google travel reviews Principal components analysis, hierarchical clustering, and k-means clustering were used in this article. K-nearest neighbors were used to classify the identified classes in the dataset. The results confirmed that, these prediction algorithms have high accuracy measures, but the clustering methodologies require further improvement. The classes identified should be checked by a domain expert for reasonableness before practical application. Because of the unlabeled data, it was not possible to test the model on new data. This model could be deployed on a small subset of customers and data could be collected on the performance of business metrics.
