Tourism Web Filtering and Analysis Using Naïve Bay with Boundary Values and Text Mining

Abstract

This paper proposes a technique to filter tourism websites and datamining to analyze keywords to improve searching on tourism websites as Search Engine Optimization (SEO). This work focuses on tourist sites or attractions, hotel or accommodation, and restaurant in the tourist provinces as a core keywords. Content in websites is retrieved from Google with queries of 11 Thai famous tourism provinces. From all retrieved results, filtering using Naïve Bayes algorithm with Boundary Values is performed to detect only relevant content as 6,171 filtered websites (66.55%) from 9,273 retrieved websites. From keyword analysis method, we compared three methods including (1) keywords from Apriori algorithm, (2) keywords from frequent terms within websites, and (3) keywords by frequency of terms from the ontology. The experiment results are conclusive that keywords from frequent terms within websites performed best. The keywords are usable to customize the websites to improve search ranking.