UTILIZING SOCIAL MEDIA TEXT MINING AND MACHINE LEARNING TO PROMOTE VIRAL TOURISM POSTING IN INDONESIA SOCIAL NETWORKS

• Main Author: Palumian (29016015), Yonathan
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/31766
• Institution: Institut Teknologi Bandung

Tourism has nowadays become a popular activity in Indonesia and its business <br /> <br /> growth has shown a promising advantage for the society. It is inevitable that tourism <br /> <br /> business has also been supported by the advancement of internet technology <br /> <br /> including social media that helps the business actors put what they can offer in the <br /> <br /> social networks. Aligned with increasing social media population in Indonesia, the <br /> <br /> threads of tourism content have also been recognized in daily basis activity in the <br /> <br /> social network such as Facebook, Instagram, and Twitter. This research employs <br /> <br /> social media text mining to gather knowledge on how people perceive popular <br /> <br /> tourism destination in Indonesia. In addition, this research adopts the STEPPS viral <br /> <br /> characteristics that are social currency, trigger, emotion, public, practical value, and <br /> <br /> story to formulate the viral promotion on tourism content in social networks. One <br /> <br /> of the machine learning techniques, decision tree learning, is also utilized to <br /> <br /> construct the viral tourism model as well as the predictive analytics for tourism <br /> <br /> content virality. The major finding of this study is that tourism content should <br /> <br /> strengthen the public and the social currency characteristic to gain the higher <br /> <br /> probability of shares or re-shares behavior then it can go viral in on social media. <br /> <br /> This research also recommends that virality social media such as Facebook, <br /> <br /> Instagram, and Twitter can be a convincing media to promote small and medium <br /> <br /> enterprises with low lost and to aim the millennial market segment that depends on <br /> <br /> social media based information search.