TOPIC GROUPING BASED ON DESCRIPTION TEXT IN MICROSOFT RESEARCH VIDEO DESCRIPTION CORPUS DATA USING FASTTEXT, PCA AND K-MEANS CLUSTERING

This research groups topics of the Microsoft Research Video Description Corpus (MRVDC) based on text descriptions Indonesian language dataset. The is a video dataset developed by Research, which contains paraphrased event expressions in English and other languages. results grouping these show how patterns similarity interrelationships between from different data, will be useful for topic-based retrieval. topic process using fastText as word embedding, PCA features reduction method K-means clustering method. experiment 1959 videos with 43753 to vary number k with/without result that optimal 180 silhouette coefficient 0.123115.