CD Skripsi
Analisis Topic Modelling Tentang Pemindahan Ibu Kota Negara Indonesia Pada Media Sosial Twitter Menggunakan Metode Latent Dirichlet Allocation
The discourse on relocating Indonesia's capital city has been a recurring conversation during various presidential administrations. Moving the capital city is considered a strategic move to reduce the burden on Jakarta and Java Island as a whole. The conversation has also taken place on various social media, including twitter. The large number of tweets related to the relocation of the capital city creates complexity for the public to understand the topics that are often the focus of conversations about the relocation of the capital city. Therefore, this research will model topic modelling using the latent dirichlet allocation method. The data used is 5000 tweets. The latent dirichlet allocation model is built using coherence score to determine the optimal number of topics, with experiments on 10, 20, and 30 topics. The results of this study obtained five topics with the highest coherence score value of 0.572237. Firstly, the topics discussed is about the obstacles in the construction of a Gas Power Plant (PLTG). Second, support for the construction of the new capital city infrastructure in Jokowi's leadership era. Third, the construction of a soccer training center under the leadership of Erick Thohir and PSSI. Fourth, the relocation of the new capital city. Fifth, government policy in opening job vacancies and investment in the new capital city.
Keywords : Coherence Score, Latent Dirichlet Allocation, Capital City Relocation, Topic modelling.
Tidak tersedia versi lain