Akademik Veri Yönetim Sistemi
International Journal of Multidisciplinary and Current Educational Research (IJMCER), cilt.3, sa.3, ss.361-371, 2021 (Hakemli Dergi)
ABSTRACT : The aim of this study was to analyze the “Black Friday-Kara Cuma” perception of Twitter blog site users via the method of sentiment analysis. A total of 491 tweets posted between 03 and 25 November 2018 from Turkey as location was subjected to analysis as data. Data were collected using the Java-based Twitter4J library. These data had pre-processed with Zemberek NLP natural language processing library. Sentiment analysis was performed based on pre-processed data. “Türkçe Duygu Sözlüğü V1” was used as a lexicon for this purpose. In conclusion, the data used was evaluated as 54% negative, 42% positive and 4% neutral perception. In addition, when examined in terms of their reasons, the tweets related to “Black Friday” cluster around certain headings to form groupings. This study emphasizes the effect of socio-cultural factors on behavior patterns in online shopping platforms.
KEYWORDS: Sentiment Analysis, Sentiment Lexicon, Text Mining, Twitter4J, Zemberek NLP.