Customer Interaction of OTOP-SMEs Products on Social Media using Data Mining Technique

Abstract

This study investigated the association between the time of day for posting to sell products and the number of users interested in the post, in a Facebook group. The data were obtained for a specific Facebook group, OTOP-SMEs Surat Thani, and comprised 1,663 records from 25 March 2020 to 25 March 2021. Data mining with FP-Growth algorithm and Python programming with the Facebook scraper library were used for data pre-processing and data analysis. The results show that the time of day with most posting was from 09:01 am to 12:00 pm (22.36%), that with the posts most liked was from 09:01 am to 12:00 pm (24.76%), and that receiving the most comments was from 00:01 pm to 03:00 pm (27.03%). Association rules were mined with the FP-Growth algorithm. There were 5 association rules extracted, such as (1) if a post is created from 00:01 pm to 03:00 pm and it is commented on, it is liked with 100% confidence; and (2) if a post is created from 06:01 pm to 09:00 pm and is commented on, it is liked with 100% confidence. The results from this study can be applied in studying customer behaviors and used to plan posting for sales of OTOP and SMEs products via Facebook groups.