Social media sentiment analysis of trending topic cigarette price issue
Aldi Andalan 1  
,  
 
 
More details
Hide details
1
Center for Health Economics and Policy Studies, Indonesia
2
Universitas Indonesia, Indonesia
Publication date: 2018-03-01
 
Tob. Induc. Dis. 2018;16(Suppl 1):A693
KEYWORDS
WCTOH
 
TOPICS
Download abstract book (PDF)

ABSTRACT
Background:
Cigarette price in Indonesia was increasingly booming in 2016. It all started from a controversial survey that conducted by CHEPS UI about public opinion and support toward cigarette price increase. The survey shows that 72 percent smokers agree to increase cigarette price until IDR (Indonesian Rupiah) 50.000. Nevertheless, this survey creates diverse responds throughout various media, TV channels, newspapers, and online media such as twitter. Hence, this research curious and want the answer from how twitter social media users responds to this issue from time to time and which words that frequently appear to the social media.

Methods:
This research use sentiment analysis and word cloud to analyze tweet texts using Python 2.7 and R It use convenience sampling (availability sampling), with 12.349 tweets are obtained. The time frame of data is from 24th June 2016 until 28th January 2017.

Results:
In sentiment analysis we define sentiments into three categories of response to the cigarette price increase. Three categories are negative (opposing stance), positive (supporting stance), and neutral (can be canceled each other or none of them). Positive stance includes lexicon of keywords, such as health, household concern, national concern, politics, convenience, support, beliefs, and public figure.Negative stance includes keywords such as beliefs, unrest, uncertainty or rumor, farmer, brands, propaganda, ineffective, alternative, opposition, persuasive, tobacco build law, and culture. Besides the most obvious keywords relating to the issues (tobacco,cigarette, price), the economic issues relating to the immediate fate of tobacco farmer is one of the primary concern of netizens.


[Histogram of sentiment analysis]



Conclusions:
The result shows that negative sentiments still overwhelm the positive ones. Tweets tend to very booming on August and September 2016 and the issues slowing again after that. Words that frequently occur despite cigarette price and tobacco is farmer.

eISSN:1617-9625