Case Study of Trend Mining in Transportation Research Record Articles

Abstract

This study employs two topic models to perform trend mining on an abundance of textual data to determine trends in research topics from immense collections of unstructured documents over the years. This study collected data from the titles and abstracts of the papers published in Transportation Research Record: Journal of the Transportation Research Board, since 1974. The content of these papers was ideal for examining research trends in various fields of research because it contains large textual data. In previous studies, exploratory analysis tools such as text mining were used to provide descriptive information about the data. However, this method does not provide researchers with quantifications of the topics and their correlations. Furthermore, the contents examined in this study are largely unstructured, and therefore they require faster machine learning algorithms to decipher them. For these reasons, the research team chose to employ two topic modeling tools, latent Dirichlet allocation and structural topic model, to perform trend mining. This analysis succeeded in extracting 20 main topics, identified by keywords, from the data. The research team also developed two interactive topic model visualization tools that can be used to extract topics from journal titles and abstracts, respectively. The findings from this study provide researchers with a further understanding of research patterns within ever-evolving area of transportation engineering studies.

Exploratory Analysis of Unmanned Aircraft Sightings using Text Mining

“…LDA visualization can mitigate the current research gap. The opensource R package LDAvis package was used to develop interactive LDA models ( 20 , 23 ). As a part of this study, an interactive webtool has been developed using the narrative texts used in 7,400 reports.…”

“…This study applied the algorithms used in Girolami and Kabán to perform the analysis ( 18 ). Several recent transportation studies applied LDA in determining topics from unstructured textual contents ( 20 – 22 ). The current framework is based on the following assumptions:…”

Systematic Analysis of the Literature Addressing the Use of Machine Learning Techniques in Transportation—A Methodology and Its Application

“…Aiming to focus on terms associated with the essence of each article, the search was limited to the title, abstract, and keywords and to articles dated from 2018 until 3 May 2022. Although several previous works attempting to analyze topics dealt with in the transportation domain focused on the most prominent transport-related journals [7,8], our initial search was not limited to specific journals. As our work aims to explore the ML techniques used for better understanding transport-related problems, we assumed that a substantial amount of them might be published in journals that are methodologyoriented rather than topic-oriented.…”

Topic Models from Crash Narrative Reports of Motorcycle Crash Causation Study

“…Text mining methods, a very effective tool in exploring unstructured textual data, can identify trends, insights, and anomalies in complex textual data with limited efforts. NLP tools such as text mining have been widely used in transportation research (17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29). Many transportation studies applied different text mining tools to explore insights from unstructured crash narrative data (30)(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42).…”