Frequent Patterns Based Word Network: What Can We Obtain from the Tourism Blogs?

摘要

In this work, we present a method to extract interesting information for a specific reader from massive tourism blog data. To this end, we first introduce the web crawler tool to obtain blog contents from the web and divide them into semantic word segments. Then, we use the frequent pattern mining method to discover the useful frequent 1- and 2-itemset between words after necessary data cleaning. Third, we visualize all the word correlations with a word network. Finally, we propose a local information search method based on the max-confidence measurement that enables the blog readers to specify an interesting topic word to find the relevant contents. We illustrate the benefits of this approach by applying it to a Chinese online tourism blog dataset.