The first step in this analysis was to collect publication records for urban planning academics. Current planning academics with Google Scholar Citation Profiles were used as the source of these data. The suitability of Google Scholar Citations (GS) data or urban planning academics has been discussed by Sanchez (2017), and provides extensive coverage of planning scholarship. Currently 54 percent (598 out of 1,109) of planning faculty in U.S. and Canada have profiles and their publications have accumulated over 75 percent of total citations (as of June 2018) within urban planning. The second step was to identify the key topics based on word frequencies and terms from publication titles. Thematic analysis was used for this purpose and produced 30 key topics (see Table 1). The set of publication titles obtained from GS Citation Profiles were then labelled using these topics with a Support Vector Machine (SVM). Based on the classification of all publication titles, citation frequency for the resulting multi-label themes were then analyzed. Three metrics were used to compare groups of publications: 1) the total number of publications, 2) the total number of citations, and 3) the average number of citations per publication per year. The number of total citations for a topic was normalized by the total number of publications and the year published. This is important because topics with greater publication activity would most likely have the largest number of citations, and older publications have had a longer amount of time to accumulate citations.
Table 1. Nodes from thematic analysis (in alphabetical order)
Using a Support Vector Machine (SVM), each of the nearly 15,000 urban planning publications was classified using the 30 labels identified (see Table 1). The process does not simply classify text (publication title) based on a single label (e.g., “transportation” or “housing”), instead uses multiple labels for more effective topic specification. This accounts for the multi-topic nature of publications (and titles) that often reflect a subject, method, and results within a publication. For example, a publication about integrating transportation and land-use planning should not likely be classified as being primarily about transportation or primarily about land-use. Therefore, a multi-label SVM assigns two labels, “transportation” and “land-use”.
Figure 1 shows the resulting distribution of labels across each of the 30 themes. Each row shows the proportions of co-labeling for each theme and the size of the dot is the relative number for that topic. The first row shows that “analysis” was frequently co-labeled with nearly all themes with the exceptions of “governance” and “engagement.” “Economic” and “environment” were also frequently co-labeled across topics. Other particular cases that are noteworthy are “housing” and “land use”, “planning” and “engagement”, “policy” and “governance”, “policy” and “local”, and “social” and “health”. The diagonal values have been removed because they represent a topic labelled by itself.
Table 2 shows that key themes in urban planning scholarship are related to economics, analysis, environment, transportation, regional, social, planning and urban. These represent the core components underlying scale (urban and regional), systems (economic, environment, social, mobility), and methods (planning and analysis). Of the 30 unique topics that were generated, analysis, environment, economic, and transportation each had over 100,000 total citations.
Figure 1. Co-labelling Distribution Across Publication Themes
Table 2. Topic frequency (in single or multiple topics)
While being a snapshot of scholarly activity, this analysis draws from approximately 50 years of publication activity from current urban planning faculty in the U.S. and Canada to highlight topics that have received the most attention in terms of publications and citations. It is likely that an analysis using more data such as from abstracts or full-text would lead to more finely grained groupings of topics because using more words could produce more nuanced labels. The analysis shows which general topics have been of most interest to planning academics (number of publications) and those of planning and related fields (number of citations).
Your comments are greatly appreciated (email@example.com).
 The publication titles were retrieved during December 2018