Analysis On Internet Pattern of Youtube Browsing in Indonesia Using Web Crawling and Unsupervised Learning (Analisis Pola Minat Tayangan Youtube DI Indonesia dengan Web Crawling dan Supervised Learning)

nfn Nofriani (BPS Provinsi Bengkulu (Statistics of Bengkulu Province))

Jurnal IPTEK-KOM (Jurnal Ilmu Pengetahuan dan Teknologi Komunikasi)
Vol 20, No 2 (2018): JURNAL IPTEK-KOM (JURNAL ILMU PENGETAHUAN DAN TEKNOLOGI KOMUNIKASI)

YouTube is a popular video sharing website, specifically in Indonesia. Every day, in every country, the list of trending videos is updated on YouTube’s Trending page. The data of trending videos can be used for information exploration, such as analysis on the pattern of interests of YouTube browsing. This research aims to grab and analyse the metadata of trending videos to generate a classifier model and statistics of trending YouTube videos in Indonesia. The data is grabbed from YouTube’s Trending page using Scraper and Screaming Frog SEO Spider tools, every day for 10 consecutive days. The data is later classified into video categories. The approach used for this purpose is rule-based classification using J48 tree algorithm and TF-IDF filter. The result of this research shows that videos about people, blogs, sports, news, politics, comedy, entertainment and music are what interest the people in Indonesia the most.