Topic: Sentiment Analysis / Opinion Mining
Research object: Tourism place in Indonesia / International
Achievement target: place rating, positive/negative, trends/hot places, recommendation
- Tweets crawling
- Data preprocessing
- Naive Bayes
- SentiWordNet will be used for Tweets crawling based on hashtag or tweet/retweet amount
- Data preprocessing technique is focused using text mining technique
- Naive Bayes algorithm is used for the classification of the opinion from visitors
- Result of the classification process will be presented in some charts that will be displayed on the website
Future work: Data source is not only from Twitter but also other social media. The combination of Twitter, Facebook and some websites might improve the reliability of the recommendation.
Tourism industry is one of important sector that currently developed and promoted by government in many countries. The government perform a big effort in promoting the tourism industry in their country. One of the promotion example is by defining tourism promotion slogan i.e. Tourism Malaysia, Your Singapore, Visit Indonesia, Wow Philippines, Amazing Thailand, etc., to support the tourism industry promotion. This is happened because of tourism industry can provide a big benefit for the country especially for the economic and social growth. Moreover, the trend of international tourism is increasing along with more free-visa destination countries. It may provide more significant benefit to the country with more international tourist.
In this situation, the World Wide Web plays a large role in the domain of tourism activity. In the last decade, there are a lot of people that book the accommodation for the travelling via online. It is because, checking on the internet is less time consuming, cheaper and they have the possibility to get detailed information about facilities and the location of hotels. Moreover, by checking the information on the internet, people may also check the opinions and experiences published by the other travelers when planning their own vacations. Some of website like TripAdvisor, LonelyPlanet, Booking, or even the official government tourism website can help those people in planning their vacations.
Here, we see that a great number of travelers generate reviews, comments and opinions about their travelling experiences. Moreover, travelers tend to tell stories about their experiences when writing reviews about tourism activity. However, these stories are more likely to have longer and more complex sentences, which often include features in them that are mentioned multiple times. Reviewers also usually mention objects that do not correspond to attributes or components of the reviewed product and use many different and complex expressions to refer to the things that we actually consider as aspects. Finally, a considerable number of sentences do not contain opinions. Sometimes, the quality of the information is a mixture between good and bad. That is why finding important information relevant to the target needs has become increasingly significant.
As the World Wide Web has developed, considerable decision making power over the consumption of discretionary products like tourism has been transferred from suppliers to consumers. The need to improve market intelligence and market research for private and public tourism organizations become a real requirement to provide a major sources of information for tourist. To overcome this problem some of research is done in tourism area. However, most of them are focused on product review and do not focus on the tourism area itself. Here, we try to conduct a research that focus on the tourism area by analyzing the travel trends that currently exists.
To perform those function, we implement sentiment analysis technique in this research. Sentiment analysis is the study of sentiments, opinion mining, emotion, etc., which expressed in text. We know that people usually use this social media to express their feeling or tell their opinion. In this study, we discuss about the opinion mining for tourist attraction by using twitter as a media to get the data. Twitter is one of the popular microblogging website. On Twitter, a great number of travelers generate tweets and hashtags related to the place visited. With this data, we can perform sentiment analysis technique and provide the result of tourism trends that currently exists. Moreover, it may also helping the tourists to get the recommendation and helping the place to develop their facility by getting feedback the users’ experiences in twitter. Finally, it can provide a major sources of information for tourist.