[npteljes]

I think a lot of data collection is done a bit like research is. The purposes are unclear, but the sea of data might become useful later on. Time might reveal trends, certain now-insignificant people might become significant, new data mining tools might be discovered that highlight something useful, things like this.