Ask HN: can I teach myself the skills needed to succeed in NLP/IR/PA/DM?

3 pointsby brosephius16 anni fa2 comments

where NLP = natural language processing, IR = information retrieval, PA = predictive analytics, and DM = data mining (had to fit the title in 80 chars :P)

for example: http://www.recordedfuture.com

their jobs page makes it sound like they want PhD geniuses. I don't have a PhD and don't really think a PhD program is right for me. that being the case, is it possible for me to do anything meaningful in this sort of field, even on a smaller scale?

I have some textbooks and read articles and blogs on the field, but I get the impression that to do anything commercially viable it has to be serious, grad-level work, not some toy someone like myself could build.

is this the case, or are there examples of successful products in these fields built by people that weren't hardcore experts?

2 comments

waterside8116 anni fa

There was a previous thread similar to what you're asking: http://news.ycombinator.com/item?id=1535869

For an example of successful products, check out my startup http://www.repustate.com

I have a BSc in comp sci but I'm by no means an expert in NLP. My partner & I just read a lot of academic papers and read up on the ideas we needed.

pedalpete16 anni fa

There are people who would ask the same thing about any programming field. I personally believe that with the tools and community available to you, it isn't necessary to have a PhD. But then again, I don't have one, and I'm just getting into learning NLP. I've done some data mining in the past, though it is what a PhD might consider trivial, it suited the purpose and solved the problem.

With respect to the link you posted. If those guys were really so brilliant at NLP, they wouldn't need you to type your query so specifically in 3 different search boxes. NLP should be able to figure that out for you.

I guess it depends on what you are looking to do. With my current project, I'm trying to extract meaning from tweets. I'm hoping to be able to get it to the point where I can put most tweets into four or five buckets of general category (self-promotion, making plans, sharing links, congratulatory). Once I have that proof of concept, then I'll look at either continuing myself, or getting help from somebody with more experience.

If I were you, I'd take the first few steps yourself and then see how you feel about it. That way you'll also be more knowledgeable if you decide to look for somebody with more experience.