Autonomous Vehicle Review
Several studies have used traditional methods established in the past decades such as systematic literature review, research thematic analysis, and bibliometrics to summarize the vast research on AVs into comprehensive and systematic forms to provide state-of-the-art information about the field of research. Nevertheless, due to the large number of documents on AVs, it is difficult to manually capture the themes in them. This study goes beyond those conducted in the previous literature by employing a novel unsupervised machine learning approach to discover interesting themes and insights from a plethora of academic journal article abstracts which can be useful for potential and current researchers in the transportation field. By exploring the temporal changes in topics, the study also identified the emerging themes which are gaining traction in AV research. The technique applied also helped identify the issues concerning AVs that are of most interest to both developing and developed countries. Comprehending and describing the changes in the overall intellectual structure of AV research over the years is likely to add significant value to the literature by identifying potentially interesting topical areas for academicians and practitioners in the transportation field.
The interactive plot that visualizes the 20 topics and the clusters in which they belong is found below. In the figure below, each topic (the apricot colored cells) is represented by a group of words that suggests the theme of the topic. To see the exemplar documents associated with each topic, please click on the topic.