Read This

To obtain insight from big data, researchers and practitioners in the tourism industry have applied different analysis techniques. Four most used techniques are summarized below:

Web crawling. Online textual data are typically extracted from social media sites via web crawling. A WebCrawler is a tool (or also called a spider) that is implemented automatically and iteratively downloading web pages, and collecting information from the web hypertext markup language (HTML). Web crawlers can also collect information from tourist websites. There are so many web crawling tools available (some are free; others are not). Feel free to read more about that individually if you are interested.

Data cleaning. Extracting big data from the web does not necessarily mean that the data are directly usable. Data cleaning processes are required to make the data easy to use for modeling and analysis. A data scientist spends a lot of time daily cleaning and checking the data for the next-step analysis.

Tokenization. It consists of breaking-up long sentences into words or significant phrases (i.e., tokens). This facilitates the identification of key phrases and patterns. Different words can be assigned to the same root as one token (e.g., satisfy, satisfying, satisfied) to facilitate modeling.

Sentiment Analysis is a method used to identify individuals' sentiments towards tourism services by categorizing textual data into different types of sentiments (positive vs negative; happy vs sad, etc.). Sentiment analysis of online and offline data can complement traditional qualitative methods, e.g., interviews and stakeholder analysis.