Fig 1.
Can–ALE walkability score for Victoria, BC.
Can–ALE scores are assigned by dissemination area. Dark green shows most walkable while dark red shows least walkable. (generated through RStudio [31] Version 1.2 from rstudio.com).
Fig 2.
ALF–Score utilizes various GIS features such as road network structure, POI as well as features derived from road networks such as various centrality measures and road embedding. GLEPO’s linear extension of user opinions that produces a global view of relative user opinions is then aligned with the GIS features as an input of our supervised machine learning processes. Walkability estimates that are produced through trained models will have a high spatial resolution, are representative of user opinion and provide a better insight of different regions and neighbourhoods. (Figure drawn by the authors).
Fig 3.
ALF–Score online crowd–sourcing platform.
Our interactive web–based data collection platform on left (map generated by OpenStreetMap, screenshot taken from alfosool.com) has been deployed with road data from various cities in Canada. Displayed here shows the city of St. John’s, NL. Total 1050 (center1) rankings were received from participants showing a well distributed data collection. Some locations were ranked multiple times (right2) by the same or various participants. We can observe in our collected data the maximum number of conflicts is 3 per location with a very little occurrence. (Maps 1 and 2 are generated through RStudio [31] Version 1.2 using mapview package from rstudio.com with OpenStreetMap used as its base map).
Fig 4.
Three examples of some of the artificial networks created to represent various possible scenarios of user submissions and conflicts.
Left: graph of 2 submissions (each with 5 nodes) containing 2 anchor nodes ‘4’ and ‘2’ forming a loop. Center: graph of 3 submissions with 4 anchor nodes. Right: graph of 2 submissions containing 2 anchor nodes ‘1’ and ‘5’ falling on two extreme ends. (Figures drawn by the authors).
Table 1.
Computational complexity of various algorithms in the pipeline.
Fig 5.
Comparison between GLEPO, ALF–Score and Can–ALE for Great Eastern Ave. in St. John’s, NL.
Left: Can–ALE, Center: GLEPO, Right: ALF–Score. A score variation ranging between 16–81 is observed within GLEPO which provides a user–level insight about how users actually perceive their neighbourhood in their own opinion, as opposed to what Can–ALE suggests of a neighbourhood to its users. Furthermore, we can observe a variation ranging between 20–70 in ALF–Score which provides a high spatial resolution rankings that are well refined and specific to each point as opposed to a single score for an entire area as observed in Can–ALE. Note: Can–ALE colors are slightly dimmed due to the adjusted opacity/transparency to better visualize the street overlay. (Maps generated through RStudio [31] Version 1.2 using mapview package from rstudio.com with OpenStreetMap used as its base map).
Table 2.
Exploration of various machine learning techniques and feature combinations with their top performing accuracy.
Fig 6.
A comparison between Can–ALE and ALF–Score for the city of St. John’s, NL.
Can–ALE (left) and ALF–Score (right). Dark green is most walkable, dark red is least walkable. Can–ALE’s scores may be rendered useless due to its over generalization. (Maps generated through RStudio [31] Version 1.2 using mapview package from rstudio.com with OpenStreetMap used as its base map).
Fig 7.
Examples of where ALF–Score and Can–ALE do and do not agree.
Each DA polygon is represented by a single Can–ALE value while each circle represents an ALF–Score. Left: Downtown St. John’s, NL, shows a strong agreement among the two measures. Right: Signal Hill region, St. John’s, NL, shows a strong disagreement between the two measures. Note: Can–ALE is represented as an overlay with a small opacity/alpha and is visualized with a significantly lighter colors due to this transparency. Legends and border lines represent the actual colors. (Maps generated through RStudio [31] Version 1.2 using mapview package from rstudio.com with OpenStreetMap used as its base map).