Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Show HN: HackerNews but for research papers

papertalk.xyz

Post reply on HN
Print view View original

201–209 of 209 posts

Re: Show HN: HackerNews but for research papers

#201

Post by jesterson »

mustang-med wrote: post #192

I'm a career scientist and here's why I suspect this won't take off: (1) A lot of research is trash and not reproducible. Over the last 5 years I've found it's usually not even worth my time to read half the papers I read unless its in Nature or a top journal. Even then the ROI is minimal. (2) It takes quite a bit of knowledge to appreciate and read a paper. In grad school, it took 4-6 hours of work to go through a p…

May I hijack the topic and ask you a somewhat related question?

Given the understandably overpolluted with garbage world of scientific research, how would look for some studies with reasonable credibility?

It's my understanding that top journal or sorts sources are likely infused with money from large corporations pushing studies for their favour. At same time smaller sources are full of crap as you noticed.

Is there a way to clean garbage from gems?

Re: Show HN: HackerNews but for research papers

#202

Post by casparvitch »

drekipus wrote: post #200
mustang-med wrote: post #192

I'm a career scientist and here's why I suspect this won't take off: (1) A lot of research is trash and not reproducible. Over the last 5 years I've found it's usually not even worth my time to read half the papers I read unless its in Nature or a top journal. Even then the ROI is minimal. (2) It takes quite a bit of knowledge to appreciate and read a paper. In grad school, it took 4-6 hours of work to go through a p…

I think the argument is that it's not for career scientists. I gave up wanting to get into academia once I found how awful the working conditions are. Now I'm in a more comfortable life, I'd love to have access (ie: visibility) on recent developments and research. I'm reminded of this story: https://spectrum.ieee.org/coordinated-robotics-winner-nasa-s...

If the return isn't there for career scientists, and it takes them 4-6hrs to read a paper properly, then can non-specialists really get much out of the process?

Re: Show HN: HackerNews but for research papers

#203

Post by casparvitch »

jesterson wrote: post #201
mustang-med wrote: post #192

I'm a career scientist and here's why I suspect this won't take off: (1) A lot of research is trash and not reproducible. Over the last 5 years I've found it's usually not even worth my time to read half the papers I read unless its in Nature or a top journal. Even then the ROI is minimal. (2) It takes quite a bit of knowledge to appreciate and read a paper. In grad school, it took 4-6 hours of work to go through a p…

May I hijack the topic and ask you a somewhat related question? Given the understandably overpolluted with garbage world of scientific research, how would look for some studies with reasonable credibility? It's my understanding that top journal or sorts sources are likely infused with money from large corporations pushing studies for their favour. At same time smaller sources are full of crap as you noticed. Is there…

Top journals have good research in them, they cannot be bought like that. The downside is that they get caught in the hype, and only publish articles likely to be more glossy.

The way to clean garbage from gems is to spend half a decade plus in grad school (literally).

If you just want to enjoy reading about the latest & greatest in science, then Science & Nature are good reading. But for any real perspective you need to read non-academic writing from those researchers willing to exert that effort.

Re: Show HN: HackerNews but for research papers

#204

Post by niemandhier »

mustang-med wrote: post #192

I'm a career scientist and here's why I suspect this won't take off: (1) A lot of research is trash and not reproducible. Over the last 5 years I've found it's usually not even worth my time to read half the papers I read unless its in Nature or a top journal. Even then the ROI is minimal. (2) It takes quite a bit of knowledge to appreciate and read a paper. In grad school, it took 4-6 hours of work to go through a p…

People differ, fields differ. I am a professional scientist too, when I go for a deep dive into a topic, I read 5-6 papers a day and take notes.

I stopped reading comics and started to just read interesting papers from adjacent fields a few years ago. Nature and Science are nice examples, most people I know treat these as entertainment magazines ( unless of course they get a paper in, than it’s the ist important thing ever).

Quality of publications varies a lot, but PNAS, Cell, IEEE are usually good sources and as long as one avoids purely computational studies one is usually on the safe side. Few people fake their experimental dat.

Re: Show HN: HackerNews but for research papers

#206

Post by creer »

sva_ wrote: post #112

I have a feeling that the tree structure of comments in which most of the discussion usually takes place under the first few most upvoted comments might not be the best to have scientific discourse, but also don't know what would be better. On the other hand, papers/journals themselves could be seen as a ultra-high latency social network in which replies happen in the form of papers that reference the work they're re…

For once it would be nice to be able to sort latest to oldest

Re: Show HN: HackerNews but for research papers

#207

Post by andrewp123 »

I was thinking of creating something like this but centered around posting one’s intuitions on research, instead of the papers themselves (which are hard to read, even the list of titles are hard to parse). But researchers are afraid to say anything stupid and would prefer saying nothing at all.

It’s a hard problem but could be a good platform for research discussions considering none exists today. Best of luck.

Re: Show HN: HackerNews but for research papers

#209

Post by sleno »

tkw01536 wrote: post #151
aDyslecticCrow wrote: post #113

Three thoughts. 1. If you could add a button that makes a search on Google Scholar for the paper so I could find its main publication and reference counter, that would be awesome. 2. I don't like being linked directly to the PDF; I would rather go to the Arkvix page and download from there. 3. What would the policy of publications be? Only open-access papers allowed? Are pre-prints okay, or just peer review? Could on…

I'd second this - linking directly to ` https://arxiv.org/abs/{{id} }` maybe even include the formats in the `Access Paper` box on the RHS of that interface would be a neat feature.

i added a 'source' button on the index and paper page that takes you to the link you mentioned, thanks for the feedback
Post reply on HN