Transferring knowledge from one field to another

Question

A professor once said to me: publication is about intelligent copy paste from one field to another. His claims were right in a sense that he knew what results already exist in one field and apply it to the other non-mature yet field. Ofcourse he didn't mean a literal copying.

If you are working in a cut-edge research (a new obscure area) and already familiar with another stable mature-enough area, is utilizing one's techniques to the other is a good approach for making contributions to the field?

is it count as a good contribution?
is it common in Academia? (specifically for Computer Science field)

Answer 1

Computer Science has many subfields, I remember that there was a running gag in NLP that people just took algorithms from the 1980's and pushed papers out like a treadmill.

Other fields like Statistical Machine Learning has a more math oriented approach and you do have to come up with novel stuff most of the time.

However, when you do interdisciplinary research, it is very often that many good techniques just don't get tested because many people do not know how to use them, or do not have the time to understand them very well.

But to answer your question, I think is a pretty valid practice, very good researchers have great papers where what they did was basically apply a widely known math tool to a novel problem that few people were working on.

An example that comes to mind is how Factor Analysis, a widely known CS method (1993) was just applied in 2007 for the problem of discovering Connections in Gene Networks, not only that, people were using linear regression on that just 10 years ago.

Answer 2

What is important in each scientific work is that it adds to the knowledge that already exists, i.e. there needs to be a new element. Applying a known method from another field into a new field, if this has not already been done before, can be a valid new element. In my view, this is a good contribution. Also, this happens quite a lot in research in general. For example, copula's where very populor already in Financial modelling, but only recently got attention for spatial interpolation.

People who come up with radically new ideas, e.g. Einstein, are very rare. Most of us researchers can be happy to add our little addition to the large pool of knowledge, mainly building on existing material.