Procedia Computer Science · 2021 · 39 citations · 10 references
Public AffairsCommunicationTransportation PlanningPublic OpinionTransportTransportationOpinion MiningPublic TransitPublic Transportation UseSentiment AnalysisTransportation IntegrationTransportation SystemsMass TransitPublic Communication AnalysisPublic Opinion EffectsPublic Opinion FramingTransportation Social NormsTransportation-related Pollution
The Government of DKI Jakarta published transportation integrated program named Jak Lingko to improve public transportation system in Jakarta. This program covers several types of mass transportation in Jakarta such as Mass Rapid Transit (MRT), Light Rail Transit (LRT), Electric Train (KRL Commuter Line), and Bus Transjakarta (TJ). In hope, the people will be interested to switch their modes of transportation to using public transportation considering the congestion and air pollution in Jakarta which is horrible. By utilizing data from twitter, this study tries to analyze public opinion on public transportation by classifying theirs opinions into positive & negative. Furthermore, topic modeling activites will be conducted using the Latent Dirichlet Allocation (LDA) method to find out topics that are often discussed publicly for each public transportation. It was found that the KRL is the only type of public transportation with the number of negative responses exceeding the number of positive responses. It was found that in the MRT & LRT, the community tended to mention words that had positive sentiments, whereas in KRL & TJ people tended to mention words that had negative sentiments. In the grouping of public opinion, the LDA model produced a topic of public discussion on public transportation that was different for each type of public transportation. The model can also capture public communication with the official account administration of public transportation. Prediction models obtained the lowest error with a Mean Average Percentage Error of 8.79%. By adding the public sentiment variable, it can improve the accuracy of the prediction model, but the tweets sentiment variable has no direct relationship to the variable number of public transportation users.
10
Twitter mood predicts the stock market
Johan Bollen, Huina Mao, Xiao‐Jun Zeng · Journal of Computational Science · 2011 · 5.1K citations · Full text
Predicting the Future with Social Media
Sitaram Asur, Bernardo A. Huberman · 2010 · 2.1K citations
Bing Liu, Minqing Hu, Junsheng Cheng · 2005 · 1.6K citations
The Role of Text Pre-processing in Sentiment Analysis
E. Haddi, Xiaohui Liu, Yong Shi · Procedia Computer Science · 2013 · 612 citations · Full text
A Study of Stemming Effects on Information Retrieval in Bahasa Indonesia
Fadillah Tala · 2003 · 278 citations