5: Predicting Hotel Preferences of International and Domestic Tourists Using Pointwise Mutual Information

Pointwise mutual information (PMI) algorithm for extracting and summarizing opinions of international and domestic travellers on the hotels during their stay. This chapter describes the significance of opinion mining using PMI as a big data analytic strategy in predicting the preferences of tourists across the globe.

According to Bucur (2015), opinion mining also known as sentiment analysis reviews the classification and determination of feelings or opinions denoted in text through the use of computing machines. The reviews comprise of opinions denoted in natural language similar to people but by computers. The opinion mining platform is used for classifying and extracting the reviews of hotels posted by travellers on tourism sites. The opinion mining processes the content of opinions and categorizes the reviews as positive, neutral and negative. Parashar and Sharma (2016) have stated that opinion mining involves constructing a structure to examine the opinions of travellers made in blog entries, tweets or audits, remarks, about the product, themes or strategy. Opinion mining predicts the emotion of an individual from sentences and orders them on extremity premise. Opinion mining is also a process to retrieve knowledge form users’ opinions from entity, objects, event and item. Opinion mining is a way to examine the thoughts, feelings of users about the real-world things. There are different sources from where opinions are gathered like micro blog, twitter, websites, etc. Kavitha et al. (2017) have stated that opinion mining has been developing widely in the past years mainly because of its huge set of applications and the scientific barriers it poses. Opinion mining is an interest research area because of its applications in different fields. Users’ opinions are useful for stakeholders and for public when making some decisions. Opinion mining is a way to extract data through search engines, social networks and web blogs. Because of the huge set of opinions in unstructured text format, it is not possible to overview the data manually. One essential problem is that opinions can be in any languages. The essential factor in opinion mining is to analyze and retrieve people’s feedback in order to know their sentiments.