This technique could help to break this monopoly of big data as it relies less on labelling which is even harder and more expensive for a large dataset than the

This technique could help to break this monopoly of big data as it relies less on labelling which is even harder and more expensive for a large dataset than the actual data.