Buzzer Detection

Buzzer detection as a design for the final project of the Natural Language Processing course that we have designed by utilizing twitter datasets spread during the DKI Jakarta governor election in 2018. We utilize the algorithm of SVC to perform classification training and by using RBF kernel. The results of the training accuracy obtained were 75% success. Detection is done by detecting words that have similarities with each other in terms of praise or hatred in classifying them.

Presentation Video: https://www.youtube.com/watch?v=aHR9mi5SIWw

Develop by:

• Kasimirus Derryl Odja
• Bryan Mulia
• Jasson Widiarta