Online Travel Forums Mining using Association Rules
There are a lot of online travel forums but they normally are natural language that make difficulty to extract useful knowledge. This study proposes an online travel forums mining model using association rule for discovering hidden knowledge in travel forums. The proposed model composes of 5 steps; step 1) collecting posts and comments from online travel forums, step 2) pre-processing text using natural language processing technique, step 3) measuring semantic similarity of words by using Porter's algorithm, step 4) transforming text and step 5) discovering association rule by using FP-Growth algorithm. This study found that: 1) applying domain of words for discovering association rule is suitable because it discovers rules that consist of 13 tourist attractions with confidence and support between 80% and 100% 2) there are several useful knowledges obtained from association rules.