3

Here is my background: I was a high school teacher that transitioned into data science about 4 years ago. I've been a data scientist and software engineer for the past 3-4 years and made a pretty good career out of it, and am now managing a small team of analysts, data scientists, and data engineers myself. However, I'm realizing I'd like to do research, specifically within computer science and natural language processing for the rest of my career. It's something I'm constantly drawn to and passionate about, and something I spend my weekends doing reading textbooks anyways, and I feel like I've hit my limits within industry. I'd like to actually contribute new algorithms and architectures, specifically within NLP (natural language processing).

Here is my problem: I don't have formal research experience- no published papers, no internships at research labs. My BA was in History over 10 years ago. I've done research projects for my work as a data scientist and deployed machine learning models into production as a result of different statistical tests I've performed, but I'm pretty sure the academic world does not consider this "real research". I do have a Masters in Computer Science from Syracuse (and graduated with the top GPA from the program) and an MBA from UCLA, and have many research papers I wrote for my classes, but were never submitted for publishing and have ended up as more or less repositories on my Github profile. Even though I was in the business school at UCLA, I've taken almost a minors-worth of graduate and senior-level coursework in CS, and done pretty well grade-wise. I also moonlight as an adjunct professor teaching a graduate level NLP and text analytics course at a top 25 university. But no research experience. I know I can do research, because I've never been so passionate about interested in this topic - and at worst, if I can't actually do it, I need to find out.

Here's what I need help with: I've seen lots of posts on different forums asking "is research required for a top X program, etc." This is not one of them. I know I have to pay my dues. I know people put in countless hours of sweat and toil in research labs to earn the privilege to be admitted to a PhD program. I'm not asking to cut the line because somehow I have some random industry experience. I've saved up enough money to quit my job (if I need to) and self fund myself for an internship or volunteer position at a research lab somewhere. I just don't know where to start - how does a 32 year old begin to acquire research experience? Do I just start picking professors from websites that have similar interests to mine and cold emailing them asking if they have volunteeer opportunities? Do I look for some sort of post-bac program? Do I try to find some research opportunities at work and document those projects really well?

Any perspectives or advice would be extremely appreciated. Thanks in advance.

2
  • 2
    I'm not in computer science so commenting rather than answering, but have you considered applying to some PhD programmes already, without waiting to get research experience? A summer in a lab is important for an undergraduate who might be unsure about what they want to do; less important for someone like you who has already switched careers to pursue their subject and is clearly well up to speed. – astronat Dec 7 '20 at 8:50
  • Agreed with @buffy that you’re almost certainly qualified to get a PhD position - go ahead and apply! Talking to some profs first is always a good idea, as you’ll be working closely with them for a few years. Alternatively, you might try looking elsewhere in industry where they do real cutting edge NLP research: google, Amazon, etc. amazon.science/conversational-ai-natural-language-processing – J. Tylka Dec 7 '20 at 12:55
4

Actually, you probably already have more "research experience" than many (most?) students entering a doctoral CS program in the US. I doubt that you need anything else. I'm also wondering if you have misconceptions about the purpose of doctoral education. This is where you get a "first" significant research experience and show the world that you can do that.

It isn't a place where you go to prove that you don't need to be there.

If you have an MS from a good place (yes, you do) then you have the necessary background and more to enter PhD studies in the US. Just make a few applications to some places that support the specific sorts of research you are interested in. Get accepted. Find an advisor. Discuss mutual research interests.

But, absent an advisor with "mutual research interests", the way to proceed as an independent would be to try to find some collaborative relationship with someone who (like an advisor) will work with you and already has some experience. You might be able to find some such person(s) through work contacts or visits to university departments, but it seems to me to be just spinning your wheels when it is admission to a program being what you really need.

Most US applicants to doctoral programs hold only a BS/BA and not many of them have significant research experience. And almost always that will have been guided by some professor.

Ignore your age. It isn't relevant. But make sure that you are able to spend three years (or so) with vastly reduced income.

But if you want it - Do It.

1
  • Thanks. My delay in responding was because I've been busy writing my statement of purpose and reading research papers. Your answer was the nudge I needed to pull the trigger. I do want it, so I applied to about 6-7 schools that have faculty active in my research area. I picked professors in my statement of purpose who have research papers on topics that I care about and explained how I would continue that path of research. We'll see what happens. I've put my best foot forward. – Yu Chen Dec 12 '20 at 8:32

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.