PEMODELAN TOPIK UNTUK MEDIA SOSIAL MENGGUNAKAN LATENT DIRICHLET ALLOCATION

• Main Authors: , RUSKE ILLA KENGKEN, , Prof. Dr. rer. nat. Dedi Rosadi, S.Si., M.Sc.
• Format: Theses and Dissertations NonPeerReviewed
• Published: [Yogyakarta] : Universitas Gadjah Mada 2014
• Subjects: ETD
• Online Access: https://repository.ugm.ac.id/131387/; http://etd.ugm.ac.id/index.php?mod=penelitian_detail&sub=PenelitianDetail&act=view&typ=html&buku_id=71845

The rise of social media analysis is currently providing a new requirement. We are required to conclude an opinion or argument in a document such as the enormous social media data as quickly and efficiently. Opinion obtained from us may infer a hidden key information and can be used for further analysis. Topic models is model for corpus to finding topics hidden in it. One model that will be discussed is Latent Dirichlet Allocation (LDA) probability model. Latent Dirichlet Allocation (LDA) is a probability model of textual data which can explain the correlation between the words with a hidden semantic theme in the document. Estimation of the parameters used in the model is a Bayesian method. Bayesian method is a method that provides value estimates through the posterior distribution. For this model the estimated calculation of the posterior distribution is very complex, therefore Gibbs sampling estimation is then used. In this paper, Latent Dirichlet Allocation (LDA) probability model is applied for data that have their source from one of the social media platform, Twitter. The aim is to know what dominant news are talking about on Twitter in a given period. The outcome of this topic models is a main topic of the entire public opinions which is then interpreted to be the most dominant news people talk about.