Criteria for acceptable integrative research

Question

Conventional research almost always involves the research of only one aspect, problem. But sometimes, I feel, there is need for integrative, holistic, synergistic research projects. Let's consider artificial intelligence and neural networks. Almost always the individual research is around the specific problems - e.g. about implementation of one specific logic in neural networks, implementation of specific language model in neural networks, implementation of visual recognition in neural networks and so on. But the practice invites use to consider the whole system, cognitive architecture, e.g. that can use the synergy of common language/visual/auditory/tactile semantics to complete multiple tasks. Such synergy can lead to improved results and hence - it is welcome from the Science as well. There is indeed the research about cognitive architectures but it is far, far less in amount and gaining far less support than the research of specific themes.

My question is - what considerations should be taken into account when one chooses the integrative projects as the theme for his/her master or PhD thesis. What can go bad? I have heard about specific requirements, advice for the doing the systems research, maybe more warning and more suggestions can be made about the acceptable and good integrative research.

I guess, such integrative research of end-to-end systems is actual not only in the cognitive architectures, but also in systems biology where one is eager to consider multi-scale models of full organisms of even full ecosystems. Maybe such research should be done in late career only? But what about the cases when internal drive, interests and perception of topicality make one to choose such theme for master of PhD thesis and nothing less is enough for keeping the momentum of drive and flow.

p.s. with the advent of the category theory there seems to be the right toolbox for such integrative research and so, maybe today, where so much facts and ideas are gathered, maybe today is the right time for doing good integrative research.

Answer 1

You need to make sure your project can be finished within the time limit available. The main problem beginning researchers seem to have is delimit their project such that it is possible for them to get it done.

The other challenge will be to find a structure for your text that would make it readable. If everything depends on everything, then where do you start?

Answer 2

"Acceptable" is a pretty broad term. Acceptable to who and for what? Actually, any ethical research is acceptable in a general sense.

However, since you seem to be studying for a MS, I suggest that "acceptable" primarily means "acceptable to your advisor" if that person must judge your work at the end. Take their advice and ignore others as long as you are a student - even a doctoral student. There are very few students who can leave their advisor behind, though they exist.

You make a good point in your question, however, by asking if this sort of research is best left to late in a career. There is some wisdom in studying very hard problems later, after you already have both a reputation and cohort of collaborators with whom you can share ideas and develop synergy. This is true in any kind of high value - high risk research. If you wind up with nothing doing problems that aren't "ripe" for solution, then you never get a chance to develop.

However, that doesn't mean that you have to ignore hard problems. Wise researchers keep a "problem notebook" in which they write down ideas they have for questions worth answering along with thoughts toward solution. If you have an idea about one of your listed problems, jot it down in the notebook. Carry index cards with you so that you can write down quick ideas even if your main notebook isn't with you at the time. Then transfer the notes when you get a chance. When you are bored, scan through a few of the problems to see if you can now make a bit of progress on them.

But spend most of your time as a student and in your early career working on things that can be accomplished without too much risk. And you don't have to work on things in strict sequence. Do what you must to solidify your career. Do what you can on the big questions.