Notes on Ley Manovich’s ‘Data’ text:

-Creating this representation involves three steps: understanding its boundaries, what objects you will represent and what characteristics of said objects you will include (this does not have to be done linearly). 

-A phenomenon (what we are collecting data on) is constituted by its representations and the conversations about it.  

-If we find correlations or patterns that describe only part of the data, this does not mean that our method is weak. 

-data= objects + features 

-Before we can use a computer to analyse a phenomenon, behaviour, or activity, they have to be represented as a finite set of individual data objects that have a finite number of features. 

 -Data representation is modular, and the features are encoded in such a way that we can calculate on them. + Only one format can be used for each feature.  

 -The most common format to organize data is a table and is used in every professional field. It is the way data society understands phenomena and individuals, and acts on them.  

 -Datasets are not just random collections of information- but objects structured in ways that allow them to exist within a computational medium.