<![CDATA[Opinion is one of the most important parts in decision making, in processing opinions require a thorough analysis process. Especially text-based opinion, where opinion in the form of opinions do not have a definite value limit for the input. Sentiment Analysis as a branch of knowledge from Text mining can be applied in the opinion analysis process in the form of text. Where opinions will be classified into 3 types of opinions, namely positive opinions, neutral opinions and negative opinions. This study grouped opinions from university graduated students using the SVM and NBC algorithms which in this study were divided into 3 main components, namely the input component, opinion grouping system, and output components.Opinion to be processed is data in the form of a * .csv format opinion file, which then conducts a grouping of opinions. Then the system produces output in the form of 3 types of opinions, namely, positive opinions, neutral opinions and negative opinions. In general, the accuracy results show the differences in the accuracy of each sentiment. From the test results generally shows the accuracy with the highest accuracy value in the NBC algorithm reaching 94.45 while the highest accuracy rate in the SVM algorithm reaches 75.76%.]]>
Copyrights © 2019
