Prediction Meaning of Words with Word2Vec on Whatsapp Data for Disaster Topic

Abstract

The community’s need for conditions that are safe from disasters is the basis for the problems of the system to be created. The system that will be created is able to detect the meaning of words from Whatsapp data obtained from the public. The data used is Whatsapp social media data, systematics of data processing starts from pre-processing and ends with an extended factorization matrix Word2Vec analysis, which is called Continuous Bag-of-Word (CBOW) to get the meaning of sentences as early detection of disaster locations. The system can extract 43% required Whatsapp data in the total of data processed by the system, and total accuracy on Word2Vec is 79%.