Sentiment Analytics as Part of Public Participation in the Transportation Scenario Planning

Monday, September 19: 5:30 PM - 7:00 PM
Location: Great Hall

 

Majbah Uddin

R&D Associate Staff, Oak Ridge National Laboratory

PRESENTATION DESCRIPTION

Scenario planning is a process to assess the effectiveness of a series of alternatives in the uncertain future. Such a process can help to make robust strategic project decisions based on the explorations of possible alternative scenarios. To better evaluate scenarios in transportation planning, public participation is an essential element. Traditionally transportation agencies rely on public opinion polls, focus groups, and other surveys to test public opinion, which is a method of involving public participation in the transportation planning process. However, those polls and surveys are often time-consuming, costly, and/or yield low quantity of samples. As a potential alternative, this study proposes and evaluates sentiment analytics based on public social media data. Specifically, it uses Twitter data on alternative fuel vehicles to determine whether the public attitude is positive, negative, or neutral based on natural language processing and data mining techniques. The scenarios considered are electric vehicle, hybrid vehicle, fuel cell vehicle, and zero emission vehicle. These are compiled from a recent report, by the National Academies of Sciences, Engineering, and Medicine, titled “Assessment of Technologies for Improving Light-Duty Vehicle Fuel Economy 2025-2035.” Over 3.6 million tweets from January 2010 to October 2021 are extracted using web scraping and based on keywords related to alternative fuel vehicles. Then sentiment analytics is applied to measure public attitudes towards various scenarios of alternative fuel vehicles utilizing the sentiment scores of tweets. The results show that the overall attitude of the public on alternative fuel vehicle scenarios is slightly optimistic. Particularly, yearly average sentiment scores are positive and range between 0.06 and 0.27. The analyses of social media data benefit transportation agencies by enabling them to monitor public attitudes and opinions for various planning scenarios.

SPEAKER BIOGRAPHY

Dr. Majbah Uddin is a R&D Associate Staff in the Transportation Analytics and Decision Sciences (TADS) group at the Oak Ridge National Laboratory. His main fields of research interest include transportation data and planning, travel behavior analysis, development of operations research models, and transportation safety.