USING SOCIAL MEDIA ANALYSIS TO DEVELOP STRATEGIES IN RESIDENTIAL PROPERTY BUSINESS

• Main Author: Fikraneesa
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/67620
• Institution: Institut Teknologi Bandung

Since the Covid-19 case appeared in Indonesia in March 2020, the growth in the residential property price index in Indonesia has been very limited and is expected until the fourth quarter of 2021. A slowdown in price increases is predicted for all types of houses due to the ongoing 10% VAT discount. This is directly impact the Residential Property Price Index (IHPR) in the third period of 2021 by 1.41% (yoy), lower than growth in the previous quarter of 1.49% (yoy). In the fourth period of 2021, residential property prices are predicted to maintain in limited growth of 1.19% (yoy). Meanwhile, the sales indicates that residential property sales in the third period of 2021 are still on hold which contracted 15.19% (yoy). The decline sales of residential properties mainly occurred in the small houses type. A slowdown is predicted in most of the cities surveyed. This paper is objective to give more insight for property developer to understand consumer’s sentiment from social media towards their purchasing behavior during COVID-19 pandemic, finding out the most and least popular factors, determining factors that requires improvement, and giving recommendation where nuance is due. Six keywors of purchasing a house are assessed, determined through manual observation on Twitter, that is housing credit (KPR), backlog, residential, property, shm, and shgb. Twitter entries from March 1st, 2021 to March 24rd, 2022 which include the words ‘house’ and its variations plus one of the six keywors are collected through a text mining process. As much 20% of data are labeled manually with their sentiment to be used as data training. This study shown that from a total of 7626 chats that had been processed previously, 5803 tweets had neutral sentiment, 363 tweets had a positive sentiment, and 1470 tweets had negative sentiment. These results can be interpreted that most of people are dissatisfied with the residential property sales and government regulation regarding residential property. The remaining 80% percent have a precision value, which is the model had explained 80% of the common things when purchasing a house.