Mining online user-generated information: adopting sentiment analysis to study emotions on Natural Protected Areas

Authors: Luyu Wang*, University of Florida, Angélica Almeyda Zambrano, University of Florida, Zhaohe Xu, Amazon.com, Inc
Topics: Tourism Geography
Keywords: Sentiment analysis, emotions, online sharing platforms, Grand Canyon National Park, text mining
Session Type: Paper
Day: 4/10/2018
Start / End Time: 10:00 AM / 11:40 AM
Room: Mid-City, Sheraton, 8th Floor
Presentation File: No File Uploaded


With the growing accessibility and popularity of social media platforms, tourists can easily generate massive contents on attractions, events, or services. The enormous amount of data is one of the most influential sources that carry sentiment and opinions on tourism related topics. Sentiment analysis is applied in many fields because opinions are core to almost all human activities and are crucial factors in influencing human behaviors. Thus researchers can describe and understand tourists’ opinions and emotions by using sentiment analysis tools. People’s perceptions and opinions of a destination and the choices they make are highly based on others’ reviews. To be most effective in satisfying tourists’ needs, practitioners need to track attitudes and feelings to examine whether the destination is reviewed positively or negatively by tourists. Here we study one of the most popular and most reviewed natural protected areas in the United States, the Grand Canyon National Park. Sentiment analysis is examined on social media data by RapidMiner. In this study, WORDNET 3.0, a lexical resource is employed to support text-mining and sentiment classification. The linguistic features of Twitter posts, Tripadvisor and Yelp reviews of Grand Canyon National Park are detected to analyze positive, negative or neutral sentiments polarity. This study investigates the sentiment changing pattern over months, comparing sentiment scores with the number of tweets. The spatial distribution of visitors’ emotions and sentimental differences among several attractions are examined.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login