Accepted Paper: Quantifying urban attractiveness from the distribution and density of digital footprints

My paper Quantifying urban attractiveness from the distribution and density of digital footprints (pre-editing PDF copy) co-authored with Andrea Vaccari, Alexandre Gerber, Assaf Biderman and Carlo Ratti has been accepted for publication in the International Journal of Spatial Data Infrastructures Research, pending minor revisions. I would be extremely curious to hear the feedback of practitioners and other researchers on that work. The abstract goes as follows:

Abstract. In the past, sensors networks in cities have been limited to fixed sensors, embedded in particular locations, under centralised control. Today, new applications can leverage wireless devices and use them as sensors to develop aggregated information. In this paper, we show that the emerging patterns unveiled through the analysis of large sets of aggregated digital footprints can provide novel insights into how people experience the city and into some of the drives behind these emerging patterns. This information has uses for local authorities, researchers, as well as service providers such as mobile network operators. To explore this capacity for quantifying urban attractiveness, we performed a case study using the distribution and density of digital footprints in the area of the New York City Waterfalls, a public art project of four man-made waterfalls rising from the New York Harbor. Methods to study the impact of an event of this nature are traditionally based on the collection of static information such as surveys and ticket-based people counts, which allow to generate estimates about visitors’ presence in specific areas over time. In contrast, our contribution makes use of the dynamic data that visitors generate, such as the amount and distribution of aggregate phone calls and photos taken in different areas of interest and over time. Our analysis provides novel ways to quantify the impact of the public art exhibit on the distribution of visitors and the attractiveness of points of interest in the proximity of the event.

The paper received good grades particularly for its originality and relevance to spatial data infrastructure research. The reviewers suggest several areas of improvement:

a) Comparative evaluation with tradition surveys
It seems necessary to provide a more reflective discussion of what exactly is the added value of our approach compared to traditional surveys(comparative evaluation with strength and weaknesses). Here I could take some of the argumentation developed by Carlo Ratti in his critique of Space Syntax: Space syntax: some inconsistencies. Then of course, I can compare with the official economic impact study performed on the Waterfalls. Anthony Townsend offers some thoughts on the mix of survey methods “the interplay between top-down systems of command and control versus bottom-up systems for collective action“.

Plus, I have to make an extra effort to describe that, following the literature, we conceptualized spatial attractiveness as a property of a well-defined place that has variable size. One reviewer seemed to have understood that our concept of urban attractiveness was based on a continuous field. Also, I have to improve with more details the data collection process (with reference to prior papers that go more into details with that aspect of the work).

b) Expended literature review
Indicators and centrality
There is a need of an expanded review of the current literature 2 on indicators, centrality, etc. to help understand the added value of our approach. This point is related to a). I could the examples of the indicators developed at Barcelona Ecologia, particularly in the domain of public space and mobility developed by good friend Yuji Yoshimura (blog). These practical example can be augmented by some of my consideration on Urban Attractiveness particularly, the mention to centrality, density and attractiveness:

In other words, city centers have been the focal point of citizen’s urban life. Therefore, the lack of monitoring of their attractiveness could be regarded as an immediate threat to the liveliness of their economy. Measures from urban data (e.g. land use, census, traffic data) and statistics (e.g. NYC’s Sustainable Streets Index) traditionally feed the suspicions and are less traditionally augmented with the systematic collection of people’s experience, as suggested by Marek Kozlowski’s PhD thesis Urban Design: Shaping Attractiveness of the Urban Environment with the End-Users that aims at obtaining subjective views of end-users and integrating this information into the urban design process.

There might be more relevant references in the Urbanism on Track book.

Volunteer generated information
Several important contributions on volunteer generated information are missing from the literature review. In particular of the recent special issue of GeoJournal such as Discovering points of interest from users’ map annotations and Volunteered geographic information: key questions, concepts and methods to guide emerging research and practice. There is also Matt Zook’s New Digital Geographies: Information, Communication, and Place (already mentioned in the conclusion of my thesis) and more recently How Does Software Make Space? Exploring Some Geographical Dimensions of Pervasive Computing and Software Studies by Martin Dodge, Rob Kitchin and Matt Zook. In the discussion on the quality of digital footprints there is a relevant paper on Addressing vagueness in Volunteered Geographic Information, still under review in the International Journal of Spatial Data Infrastructures Research, Special Issue GSDI-11.

It will be important to refer to the more practical works that have emerged after the submission of the paper such as Mapping the cultural buzz and Mapping the World’s Photos

c) Density estimation and Spatial interaction modeling
It would be good to link this work to the literature of density estimation (see Bernard Silverman’s Density Estimation for Statistics and Data Analysis) nd of the estimation of attractiveness in the context of spatial interaction modeling (Fotheringham and O’Kelly’s Spatial Interaction Models: Formulations and Applications). A reviewer mentions this would help to place the work on a sounder basis, and to address the kinds of “why” questions that would make the paper more than descriptive.

d) More work on figures
More attention could be paid to the figures, particularly Figure 3 could display isolines, location points of photos could appead on Figure 4 and there could be an effort to map more effectively the urban indicators such as attractiveness and PlaceRank

Relation to my thesis: This paper is part of the research on the Next Generation Digital Earth, promoted by the Vespucci Initiative. I have a lot of sympathy for their open research model of the International Journal of Spatial Data Infrastructures Research based on creative commons license. This greatly facilitate the timely exchange of ideas and feedback. The publication of this paper is important to close my thesis. The research on the NYC Waterfalls generated also generated a conference paper focused on estimating the presence of visitors of the exhibit. It will be presented at the 11th International Conference on Computers in Urban Planning and Urban Management (CUPUM)
Related to: Reviews From Latest Submitted Papers