Aspect based opinion mining of hotel reviews using apriori algorithm for frequent aspect finding
The emergence of Web 2.0 leads to the development of online review, such as product and hotel review, where the customer can freely write their opinion towards the hotel. Consumer opinion is the important source of information for people when deciding to book the hotel. However, with the increasing amount of review, it is making impossible for people to read all of the review. Additionally, it is also hard for hotel management to monitor the consumer opinion.rnrnAspect Based Opinion Mining is a study of people’s opinion and the aspect towardsrnthe entities. This thesis is a research for finding the product aspect and the opinionrnpolarity from the hotel review. There are 3 steps to perform the research: (1) Identifying frequent product aspect using apriori algorithm; (2) Identifying polarity ofrnthe review using SentiWordNet and CoreNLP; (3). The Precision and recall for predicting aspect and opinion pair is averaging 0.57 and 0.64 respectively, with the average F-Measure of 0.6, and can generate average of 75% of relevant frequent aspects. Using SentiWordNet and CoreNLP has high accuracy for predicting the polarity of the sentiment, with the average of 0.86 and 0.8 respectively.
|B02499||(Rack Thesis)||Available but not for loan - Missing|
No other version available