So I'm trying to get into computational linguistics, I'm almost finished my undergrad in mathematics, with a CS minor and some linguistics classes under my belt. I'm probably gonna try getting into a masters program eventually.
What I'm really wondering, though, is whether anybody here does work in this field? What type of work does it entail? I enjoy the theory and the actual study of it, I've recently been going through an NLP course on Coursera, but I don't know what it's like at the professional level.
How is linguistics? I know many comp sci people who chose it as an elective and are pretty disappointed by how boring it is.
Jose Lee
I actually enjoy it a lot. To be honest, I think a lot of people who study CS today aren't interested in the sciences or even math all that much, they just want to program; I'd take their opinions with a grain of salt.
That said, you might find parts of linguistics boring, there's a lot of subfields and people definitely find their preferences. I tend not to like any of the stuff that relates to sociology, which is I think a growing field but really isn't the same kind of study as the rest of linguistics, though my intro to linguistics professor seemed pretty big on it.
Samuel Campbell
compling in the industry boils down effectively to NLP/MT in a very specific domain and rarely does it deal with any of the theoretical content that the degree generally touches on. i am going to get a compling degree sometime soon, but if you enjoy the theoretical part then i recommend following through and getting a PhD.
also the only interesting places to work (I have heard) are google/MS/facebook/apple/amazon, any other place you are doing some really garbage tier programming work for some dumb shit application. i recommend if you cannot work for these places out of your masters program, you should just take a regular software engineering job.
Jonathan Sanchez
Thanks for the advice, I was teetering on whether I wanted to go for a PhD or not, just because of the time and cost involved, but from what you say it makes it seem like my best course of action.
What do you plan to do? Go into a tenure-track position or something?
Alexander Harris
If you are good you can work on shit like google translate or other translation software. I think you'll get some nice salary from it.
Dominic Wright
>google translate >linguistic
Kek. These shit are full machine learning now.
Jaxson Bennett
Just what kinds of things do you think go into machine learning? Tons of the work behind that technology is computational linguistics.
Levi Carter
>machine learning >comp linguistics I think you mean deep learning
Andrew Clark
Statistical analysis and training are things that go into parsing, which is used in translation.
Carson Morales
Exactly. They machine learning goes into linguistics, not the other way around. Linguistics would be subset of machine learning with intersection in a few other fields.
Hunter Cook
>Linguistics would be subset of machine learning with intersection in a few other fields. Please don't speak on matters in which you have zero understanding.
Parker Reyes
Linguistics would be a pipe dream without statistical learning methods. They outrank them.
Ryan Stewart
>They machine learning goes into linguistics, not the other way around. That's not how it works. They're complements of each other in the realm of computational linguistics, but machine learning has *nothing* to do with the field of linguistics itself.
Please, seriously, do not talk about subjects you've clearly never actually studied.
Jaxon Stewart
>That's not how it works. They're complements of each other in the realm of computational linguistics Right, and in the general scope of academia, they can be considered two distinct fields, as machine learning can be used entirely without application to linguistics >but machine learning has *nothing* to do with the field of linguistics itself. Exactly. Machine Learning methods are used within Lingustics as a tool, but ML itself specifically has nothing to do with Linguistics, while Linguistics relies heavily on ML. Ergo, Linguistics is the subset with intersection in some other fields and would be a pipe dream without it.
Ayden Murphy
>Machine Learning methods are used within Lingustics as a tool, but ML itself specifically has nothing to do with Linguistics, while Linguistics relies heavily on ML
lolno
Standard linguistics has absolutely no use for machine learning. Only in natural language processing, which is distinct from linguistic theory, is it used.
You do not understand linguistics, or likely machine learning, at all.
Juan Morales
>Modern computational linguistics is often a combination of studies in computer science and programming, math, particularly statistics, language structures, and natural language processing. Combined, these fields most often lead to the development of systems that can recognize speech and perform some task based on that speech. en.m.wikipedia.org/wiki/Computational_linguistics Computational linguistics is an intersection of fields which relies incredibly upon statistical learning methods aka machine learning. "Standard linguistics" is moving heavily in this direction more and more every day and much of Theoretical Linguistics concerns measurable and observable data including loudness, color, shape, amplitude, frequency and the ability to interpret those and other similar audio/visual data as language. Theory was always heavily based in hypothesis formulated based on statistical findings from the basic data points, and topics such as semantics and syntax extended from established lower level hypotheses. Gtfo off muh Veeky Forums
Easton Smith
Amazing, you used the page for computational linguistics when I said that was the only place machine learning was relevant to actual study.
Saying something like machine learning being used for data collection makes the field of linguistics a subset of it is like saying physics is a subset of measurement.
Hudson Jackson
Physics is a subset of measurement
Wyatt Reyes
k
You're confusing the tools you use with the actual subject. You probably think geometry is actually about describing the earth, or that computer science is actually about computers.
Gavin Brooks
No silly. Geometry is general measurement theory for real and theoretical shapes and objects, and computer science is applied computational theory, which is a subset of math.
Adam Johnson
this. computational linguists are computer programmers who happen to work on projects related to a language or group of languages; they don't study language as a mental faculty as in real linguistics.
