Recommendation algorithm is at the heart of personalization of contents!
Why? The answer lies in the growing importance and availability of data and speed of changes.
This of course is breaking a tradition where human knowledge and insights drive designs and decisions directly – in case of personalization, many of them are incresingly mix human manual process with data and algorithm driven process and in many other cases, it can be completely data/algorithm driven.
We should really not be too surprised about this if we stop and ask, where the human knowledge and insights come from? It based on data, many different kinds of data. When the relevant data are sufficiently available and the learning process is well understood, put human effort in between the otherwise automatable processes can only add inefficiency.
I have just run into this interesting post about music recommendation, a field rich with many different ways of doing personalization/recommendation. Here it is: Four Approaches to music recommendations.
Analytics Strategist 