FYP Abstract
In the world where technology is still advancing, online social network really help human in communication. One example of online social network application is Twitter, this application is one of most widely use. Thus, creating a new threat toward the human society. One of the threat is cyberbullying, in which this issue is difficult to be solve. Not only young people being attack by cyberbully, adults also prone to this threat. Even is this threat did not cause any physical harm, but it is surely deal a lot of metal damage. Thus, some researcher tries to come with solution to prevent the cyberbully post before it reaches the victims. The solution is to create a method that would detect the cyberbully post. There are variety of detection toward cyberbully, but all of it have a similar problem, that is it cannot detect post that use slang in it. To make matter worst, the use of words in Twitter is so dynamic that it will change in time, making detecting slang words really difficult. Therefore, this problem of slang need to be over and solve. Thus, this research implemented slang normalization into Fuzzy Logic classifier to detect cyberbullying post. In this research, it is tested the method to find out whether this method is better that other. After the results analysis, this research found out that the implementation of slang normalization in Fuzzy Logic Detection Method did improve the effectiveness 21%.