|
Internet is getting easier to be accessed, thus, information is easier to be obtained and easier in reaching each other around the world. Before making any decision, especially before consumption, besides of asking opinion of friends around, people are more likely to search from the internet. However, during the information explosion era, are all of the sources of information effective? If the answer is ‘No’, then how are we going to identify the authenticity? This question lead to the development of this research.
This research contains two main stages. The first stage is to obtain information by using text mining, they are web review and web rating from a traveling website. In the second stage, we use the obtained web review to analyze their sentiment. After sentimental analysis of the web review, we verify the result. If the review was identified as positive sentiment while the rating is also positive, it means the result matches, and it is a valid review. If the review was identified as negative sentiment while the rating is also negative, the result also matches, and it is a valid review too. If the review and result it not matches, then, it will be identified as an invalid review.
After a verification process, it is proved that the positive and negative sentimental scores of this research are consistent with the rating from the reviewer, especially in the negative sentiment analyzer. It means the system produces consistent sentimental analysis result with both the positive and negative rating from the reviewer, therefore, it is reliable.
We hope that by introducing our research method to determine whether the web review and rating are consistent, to judge whether it was valid or not. If it is able to identify the validity of the web review, they will become more convincing, the user will not be misleading by invalid or false review and fall into the trap, and then make the wrong decision, but not aware about it.
|