Treatment and harmonization of data: the right recipe

Urban analysis is a long string of information, treatment of data collection (on the ground, import of databases, crowd-sourcing…) to viewing the results integrated into a virtual city modelInformation lost not being able to recreate itself, the quality of the results is limited by the weakest link in this chain. If it is a particularly critical link, it is treatment and harmonization of data, following the phase of data collection.
My friend Piergiorgio de Ferrara likes to compare this step in a recipe. Let me borrow this tasty analogy.

 

Excerpt from the pie with leeks recipe

A recipe is a methodology. It lists the ingredients, stating their quantity and unit of measures, and details how they should be prepared before be integrated with each other.

 

Leeks purchased on the market must be quartered lengthwise, washed, and then cut into section of 2cm long before being put in a saucepan suitable for realize a pie with leeks.

Similarly, the collected raw data must be made usable in the urban analysis, transformed and eventually corrected to match the expected format. The use of code-lists specifying acceptable values of the text settings (function of the building, type of owner) has an important role in the harmonization of data. For digital settings, a mastered uncertitude is critical to the quality of the data.

Pre-prepared items (kub Maggy) to replace the missing ingredients cheaply. Just as libraries of data, collecting data for benchmarking that will be used by default if the stage of data collection did not fully provide the data needed for urban analysis.
In addition to the necessary ingredients, ingredients additives can enhance the taste of the dish (spices, truffle chips!), just like the optional data, which integrated into the processing of data will improve the accuracy and realism of the results.

Finally, without kitchen utensils adapted to the quantity and the nature of the ingredients, point of salvation. Similarly, data processing tools are essential. Be they commercial (FME) or freeware/opensource (HALE), there is something for every taste, for every budget, more or less strong and sharp. Also need to know to use: cook it as the expert in data processing must have the necessary experience, so that nothing will be left to improvisation.

Leave a Reply

Your email address will not be published. Required fields are marked *