PENULIS

Adat Muli Peranginangin, Albert Owen, Felix Emmanuel, Luis Jonatan, Winnie Lauren

TANGGAL

20 Februari 2026

ABSTRAK

The growth of the internet has resulted in the digitalization of data, which has led to the emergence of
big data opportunities. Significant amounts of digital data leave traces of what customers see, read, do,
and judge, as well as information about their interests and preferences, resulting in a large amount of
data that may be mined for learning experiences. Data mining, statistical algorithms, and machine
learning approaches are used in descriptive, predictive, and prescriptive analytics to analyze, forecast,
and optimize what is the most take effect, future trends, events, and behaviors based on various data
types. A decision support system is widely demanded in tackling this problem, especially in
understanding the interactions based on the type, and time from the Facebook post about branding data
sets. This work attempts to offer descriptive, predictive, and prescriptive analytics to determine
whether a post is worth paying for and promoting. This study is sought for deeper observations of
posts on Facebook that get a lot of interaction and loyal users by the best algorithm compared with
naive Bayes and decision tree which is using Random Forest with 90.35 % accuracy.
Keywords : Business Analytics, Random Forest, Support Vector Machine, Naive Bayes, Decision
Tree, Data Mining.

FILE YANG DAPAT DIUNDUH