| Summary: | This study explores the sentiment of road users on Twitter and provides detailed insights into how the sentiment towards road infrastructure and conditions is perceived in a social media environment. The analysis plays an important role for the government, considering that Selangor is one of the major contributors to the Malaysian economy. The study's results can be used to inform local authorities and urban planners about the current condition of road infrastructure. The sentiment of road users was conceptualized as a multilingual construct, considering both Malay and English words, using a text-mining approach. Tweets were collected through the Twitter application programming language (API) and labeled using VADER and SentiWordNet, which are two commonly used lexicon-based techniques. Several machine learning classifiers were explored, namely Support Vector Machine, Naïve Bayes, and K-Nearest Neighbor algorithms. The Support Vector Machines algorithm with VADER lexicon and 20-fold cross-validation yielded the best performing model with an accuracy of 78.67%. With the exception of the Sepang district, the number of negative polarities exceeds that of positive polarities. It could indicate that the Sepang district road is high quality and has fewer hotspot barriers. Additionally, the insights were presented in a dashboard for better understanding, as the contribution for road infrastructure sentiment analysis with classifiers. Nevertheless, including more cities and states in the study would enhance its generalizability. The findings provide insight for the target users, particularly the authorities managing and maintaining the road, and raise public awareness about Selangor's road infrastructure and conditions. © 2023, Success Culture Press. All rights reserved.
|
|---|