After reading this question regarding arXiv and finding a link to GitXiv which I was previously unaware of, I wondered, what websites/resources do people use to keep up to date on the latest research in their area?
As a data scientist, I need to keep up with the latest research in the computing, software and the data domains. And here are some things which I do to keep myself updated (apart from lurking in Gitxiv):
- Quora: Amazing trove of information. This is a place where one can get questions answered by experts and also learn about information. So, it is a great place to keep oneself updated about the advancements in the computing and the software industries and the domain of interest.
- Data and software Blogs: There are a plethora of data blogs on the internet. All of them have rich theoretical information, and some even explain real world data projects really nicely. The software blogs help me in keeping myself up-to-date about latest software practices and latest data science software too. Example
- Engineering Blogs: The reason behind listing this as a separate bullet is because they are addictive. They are excellently written, and talks about the state of the art data and software practices followed at some top companies. Some of my favourite are: Buffer Airbnb and Pinterest Bonus: This Quora question
- LinkedIn groups: LinkedIn groups are a great way to ask questions and interact with some of the best in the industry. The LinkedIn pulse is also a great way to keep oneself updated with wonderful posts from top influencers.
- Asking people doubts: Asking people specific and nice doubts would be very helpful and a nice way to interact with some of the best in the industry. A related example.
Attending the seminar and conferences is the best way. When you talk to active researchers they'll share more background about what's going on in the field, rumors, unfiltered opinions etc.
I remember in my old days in physics, we had this group of researchers who published a lot, but those who're in the field knew that it was junk. So, if you're not going to conferences it would be hard to figure out.
UPDATE: I also have a profile in Google Scholar and ResearchGate. Both recommend papers based on what papers you publish and read through these portals. I found that both produce a lot of noise, but Google Scholar hints to relevant interesting papers more often. I also look up almost always first in Scholar, so it knows well what kind of stuff I often look for.
The main thing you're probably looking for is a way to quickly weed out the "junk" that's not interesting to you in whatever field you're in.
Emails from journals listing their recently published papers are good, but even better is things like RSS feeds, which allow you to aggregate results in your feed reader of choice. Having the results of multiple journals in a single place allows you to rapidly sort titles into piles like "read", "don't read" and "maybe". But often there are so many journals which touch on your area of focus that even that can be unwieldy.
I'm holding out hope for machine-learning type recommendation engines that can learn what sort of papers I'm interested in and sort things automatically. Sort of like a Netflix/Amazon recommendation service, but for journal articles. There's none yet I can recommend whole-heartedly, but I've played around with Sparrho, and it seems to work decently. Another recommendation engine site I'm aware of is PubChase, but that's biomedical only.
Ps: Im only answering this question with the intentions of helping a mere mortal like me who takes a little long to go through a typical machine learning research paper and doesnt mind reading layman sumaries of this research papers.
To keep up to date with the latest machine learning and deep learning research, I usually go through summaries of research machine learning papers, written in layman terms from the following sources:
Again: this is what I do to keep up to date with the latest research as I tend to take long to go through a typical research paper.