Each row considers the number of pedestrians at a certain location for a specific hour and weekday and the temperature and weather conditions. Eventhough the direction of the movement of the pedestrian is supposedly accounted for by the scanner, I have not found any information about this (neither in the data documentation nor in the dataset). Again, this seems to indicate different purposes: if one is interested in how many people pass by a shop, directions is not that relevant. If one wants to know how many people are attending a specific event, it might be relevant. Interestinly, it provides data that is of no interest to the users (interview: we are not interested in data beyond the open shop hours). The interview also indicates that more context-data is added (times of sale, Black Friday, or a specific event) to make sense of the data. How is it decided which cata need to be counted automatically and which are added "as common sense". the data obviously can only tell that and how amny pedestrians moved through the scanner, but not why and for what purpose.
We don't learn much again about the actual social processes, but some considerations that were decisive. It seems data protection requirements played a role. I find it also important to note that an existing (commercial) data model was taken on by the city / the economic development department. As the controversies over the potential damage for retailers /shop owners are central for planning and transforming urban traffic, one could ask waht happens when data coming from a certain logic are then used for a different purpose.
our knowledge of the history of the particular dataset is still a bit limited. From the interview, we learn how the decision for installing the scanner was made or at least the intentions behind it. Interesting ist that there are several units involved that are at the intersection of common interest and economic interests. The unit that installed part of the scanners is linked and funded by the city of Frankfurt with a focus on enhancing retail in the city. The company that offers the scanners and had already installed one before the city "jumped on the train", is part of an investment company for commercial real estate. The company describes itself as coming from a catholic tradition and describes itself as conservative in its investment strategies. One of the central aims is to preserve the values of "lively inner cities" which, as claimed on their website, are bound to "Heimat and Identität". I don't see how this is linked to the data set / classification system itself but it seems worthwhile to consider the insterests and values claimed by those working with it. The aim of Hystreet (which was founded as part of the investment company's strategy for enhancing data analytics) is to provide retailers and urban planners with the data basis to adjust their strategies for preserving retail in inner cities. The same is true for the economic development unit. This seems to cover many of the stories of "urban data": the interests and values that led to generate ana analyze certain data are very diverse and somehow fade into the background when put together to create more representative data.
Looking at the actual dataset and its rows and colums immediately shows one difference to the bike data: weather seems to be particularly important to interprete the pedestrian data in the right way. Most probabaly, becuase such data are easy to collect in addition and in the same resolution.
It seems the interesting decisions are less within the classifications system itself, than with regard how the data are generated (interview: we decided against hanging them close to the subway entrance as it traffic is not the core question - so the SITES of data production seem important) and how they are interpreted (interview: the red light model that was understood/used in completely opposite ways).