Hacker News Reader

Comment by hack_edu | Hacker News Reader

The state of academic publishing is in a truly dreadful state. The same monopolies that created the industry hundreds of years ago have never been pushed out and innovate in the slowest way possible, aqui-murdering even worse than EA when they have to. Online journals and the promise they provide are peddled by high pressure sales pitches who insist on absurdly high prices. For many institutions, when taking usage stats, the price of online access per user often works out to several dollars per search...

This is why I've given up trying to work in the field or found a company in the sector... Anyone with an idea and need a technical cofounder? ;)

View on HN

I think that there should be something like SO (Stack Overflow) for scientific papers, with anonymous user-generated reviews and ratings of the papers, and reputation being acquired by the users, both based on papers and their reviews. I wrote 'anonymous' because it's a more heavy-weight business than posting and assessing questions about coding. To avoid the situations when scientists get angry because of bad reviews, and try to revenge on their reviewers, it should be at least in the first years anonymous. The reviews would be voluntary, and also would be rated, just like answers to question in SO. I am a scientist, and I know that reviews have very low quality because of many reasons, such system could perhaps help improving the reviews, while keeping the reviewing process automatized, and reducing its cost to the minimum. In fact, costs of storing papers are low (Arxiv), reviews are free (nobody pays to the scientists anyway, they do it for the community and reputation), and the only cost that is high is the cost of managing the reviews. But if reviews could be automatized then we have a system that is as cheap as Arxiv is, and peer-reviewed :)

In the long run I see such system emerging and being successful, and eventually replacing everything else. Needless to say it will take a lot of time though to popularize it.

What do you think? Could this be done?

Refereeing papers is an odious chore that you only do to get your back scratched in return in the future. (They call it "giving back to the community", but it's really a protection racket.) You can't replace this with a voluntary system.

Why not, if you would get reputation points for it? The idea is to review the papers which you want to read anyways, or perhaps you have already read them and you want to express your feelings about these papers, hopefully in a way useful to other viewers of the paper. Nobody tells you to do it, so you do it by heart, this surely would increase reviews quality.

Well, maybe this could work, but probably it would have to replicate the "coercive" component of the current system. In other words, by reviewing other people's papers you somehow improve the chances of having your own papers reviewed. And if somehow that system manages to also provide an incentive to do the job thoroughly, then that would be a huge improvement on top of what we have now.

Well, the coercive component is the unified reputation that you gain in this site, just like the reputation that you gain in SO. Obviously, the higher reputation you have, the more recognizable you are and more people read about what you do.

Coercion comes from the employers (tenure/promotion requirements) not from the publishers.

I think that could be a successful community, you just have the same problem as any social community, finding users. Especially ones of interest that are open to publishing anonymously and listening to anonymous reviewers.

The real tragedy is the general reference, social science, and humanities journal diaspora where the usage and demand has skyrocketed with the increased access online resources allow to lay researchers. The science journal communities are way ahead in self-sufficiency in adapting, and are increasingly open to the corrupt state of publishing and alternate community publishing platforms.

I think SO is the wrong model and that Hacker News is a better one. There's already reputation/badges/etc for the sort of academic tasks you're describing: look at any CV (eg: conference presentations, referee assignments, etc, and the obvious one is journal publications). So I don't see how replacing the old system with the old system plus usernames is an obvious improvement or a necessary evolution.

I'm way more interested in informal discussion of interesting papers and news than anything else.

As everyone else has mentioned, the big issue is getting a large community. So what would be really cool would be to figure out how to do something like HN or SO without active involvement by a large community. In my field there's a lot of discussion/review/etc of working papers and published results, but it's all scattered across different blogs and commenting systems. Plus papers build on each other and some are clear replies to other research. If you could take that sort of dispersed discussion and put it into an HN or reddit style threaded view that was available at one spot (or an SO style question and answer format, if you prefer), that would be awesome.

edit: typo

There are a few websites already trying to do this, but as with all networks, it only works if you get enough users.

Could you give me examples? Thanks!

ResearchGate (http://www.researchgate.net/) and Academia.edu (http://academia.edu/) are two examples of companies looking to build networks for academics.

Arxiv and marxiv are already working on this.

Arxiv even has a paid fellowship, but it is postdoc wages, not professional wages.

Sounds really good. Could you link me to some material about this?

Serious question: why does this need a technical solution? Researchers comprise the editorial staff for these journals, conduct the peer reviews, and submit the papers, all for free (or nearly so). They (we) universally want the papers to be accessible by other researchers (because that means the paper might get cited) and are at worst indifferent to the papers being accessible to the public at large, and by that I mean some people care a lot about that issue, some people don't care very much, but I don't know anyone who is opposed to public access.

So there's an ongoing transition now by researchers/editorial boards/etc to move to a system with lower access costs (some of which involves taking back journal ownership, some doesn't). There's some logistical overhead and fixed costs, but those are falling. It seems like in 20 years (conservatively) all new research will be available to whomever wants it no matter what Elsevier or other companies do in the meantime (the papers may not be literally free, but affordable).

So, like I said initially, I can see how some software for logistics or communication would be helpful, but the main issues don't strike me as technical.

One of the many problems with academic publishing is that it's holding on to an antiquated idea behind publishing and not progressing with the way communications are moving now. In a sense, publications are moving at 1970s speed and this hurts science. Hence the technical need to bring it to current times.

Is it a "technical issue"? No, but neither is booking your flight via phone, that worked just fine. Could it be made much better, more accessible, cheaper, and overall better via technology. Yes.

An anecdote: I finished a decent draft of a paper two weeks ago and posted it to my website an hour before submitting it to a journal. The code to replicate it is on github available to anyone who wants to look at it, and the statistical methods I use I've put in a GPL R package. I've already discussed aspects of the paper and some implementation issues with other researchers over email, presented the paper at conferences, and I think people are already writing papers that use the statistic I propose (again, despite the paper having been submitted to a journal for the first time two weeks ago). In the process of discussing the paper (in person, over email, etc) I've already essentially done two rounds of peer review without journal submission.

Maybe you could give details about what you mean by "antiquated idea behind publishing" because I'm pretty sure what I described is not 1970s speed. Now, if your point is that what I've described is not common practice, I agree completely. But that means that people need to be persuaded to use existing technical tools -- I can't see a specific step that's going to be a significant pain point (I'll grant that not everyone can/will use github, but uploading a zip archive of the paper's directory to a personal webpage is easy and is almost as good for code and data dissemination).

Other anecdotes are blogs that announce and discuss new research:

http://johnhcochrane.blogspot.com/2012/10/nber-asset-pricing...

http://johnhcochrane.blogspot.com/2012/09/woodford-at-jackso...

That is legitimate academic discourse, put out by an extremely established and credentialed economist, almost instantly. On blogger.

I think it's super awesome that you did all of that! Kudos to you, and I wish more people were doing stuff like that.

My viewpoints are limited to life sciences, where people have no idea what FTP means, let alone GIT. Also, what you did is not typical, people (at least in my field of neurobiology) have no intention of doing something like that in fear of being scooped. The antiquated idea behind publishing is the concept of doing an experiment, filtering out all the negative data, and then submitting their positive data to a journal that they think will publish their data. We've been waiting for 7 months now for a paper to get published in PNAS due to this antiquated process. 7 months to do what you did in a few minutes uploading your data online and making it public.

I'm all about this sort of openness in the research process, but it seems counter to some disciplines' cultures- aren't eg, medicine and chemistry more competitive / secretive prior publication?

You're doing it vastly different from many academics. Kudos to you!

A good start would be increased quality of interfaces and abilities to find and connect similar sources; that's all I'm really asking for. There is a huge amount of research published within the Gale-Wilson-ElSevier-JSTOR-Ebsco-Taylor&Francis vaults, properties of immense academic and economic value, yet the tools they charge to access them aren't much more advanced than Yahoo circa '99. Don't even get me started on the transition from physical-copy journal publishing where you always 'own' the copy to a model of journals-as-a-service (the ultimate DRM; they host it) that cost well into the five figure range per year where you lose all access once you let the subscriptions expire.

Meanwhile, 'search' as we know it rules the world and is the driving force enabling the entire tech industry. It wouldn't be hard to throw a little effort in building out their platforms, but there's no competition to force further development in their products.

Not sure you realize, but these companies already are throwing more than a little effort into advancing their platforms.

Elsevier in particular is spending ridiculous amounts on internal research & acquisitions for its platform (called SciVerse, http://www.hub.sciverse.com/), following the latest advancements in machine learning and NLP for data mining, etc etc.

They also host competitions for start-ups and developers to produce new apps for their platform.

These businesses may be unethical and not to everyone's liking, but they are not stupid.

For most scientific fields, the last decade's research is much less important than the next decade's research. New research is much less locked in than stuff that was written before the internet became ubiquitous, so I'm not as concerned about that issue.

Well, a technical solution is needed to make changes faster and even better, namely: 1. Provide for the papers a system of transparent reviews and ratings accessible and readable to the public. 2. Provide a recommender system and a search engine based on the reviewed and rated papers available publicly.

This would add real value to the scientific publishing and simply help the researchers to find what they need.

> Provide for the papers a system of transparent reviews... [snipped]

I don't think this is a technical issue. Journals make mistakes on which papers to publish all the time, even when the referees work hard and take the review seriously, so this is hard for committed people to do well.

Sure, but reviews are not transparent, nor accessible to the public. Why not? Wouldn't it be nice to see what other people think about the article before reading it? Currently there are citations, but that's about it.

My point is only that these are cultural issues, not technological issues; I agree that it might be useful to see referee reports.

A public review of article X by reviewer Y would often be quite different than a double-blind review of the same article by the same reviewer.

There are some arguments for both approaches, but I believe that the double-blind approach is much better.

I dont know how it is in other fields, but in physics, at least, reviews are only singly blind. The reviewer knows who the author is. Even then, it is often possible to guess at the referrer based on the comments they make. Blinding the author would be nearly impossible due to the small number of workers on most specific topics. That being said, I'm not sure why journals don't at least try (which should benefit newcomers to an area).

I don't think any journal reviews attempt to be double-blind. I could be wrong, but I've never heard of it in my field. In smaller fields, even single-blind (the author not knowing the identity of the reviewer) is difficult, the other way around would be impossible.

Yes, I actually do:) Me: Neuroscience PhD at Columbia w/ some a technical background but not a coder.I'm working on going up against academic publishing right now. I have a YC application that is currently being reviewed and I'm in need of a technical cofounder.

Project: Very briefly, I'm working on collecting raw data (positive AND negative), indexing it, and making relationships between datasets. For example, my neuroscience study has relationships to cancer biology, so the data presented to the user would include results from both.

Re: Publishing, I'm going to provide raw data to scientists and create a venue for them to "blog" about these datasets. I think science needs to move incrementally (at internet speed) and get away from taking 3 years to produce a study and publishing a big paper of only positive data. Also, I'd like for the "peer review" portion to be more transparent, so that we can avoid the insane bias that goes into peer review.

Goal 2: Create "big data" by combining small academic science results into a big ass database. Meta data could be the new preliminary data for academic grants.

So yeah, if anyone is interested in playing with this, let me know. klg2142 @t columbia.edu

Aqui-murdering is a consequence of our financial system. When some big company can get bank money at far less than other companies, let's say 3% just as a hypothetical example, they can buy any business with a 6% cap rate and run it poorly so it only makes a 4% cap and still make a profit. This happens all the time. Meanwhile the entrepreneur cannot get money for anywhere near that and has to actually create real value instead of performing financial arbitrage as these companies with access to the cheap money can.

There was a time when I would've been on board. But I think that time has passed. I have a wife, a new baby, and a nice day job that allows me to enjoy both.

A couple years ago, I spent many nights and weekends working on the academic publishing problem. The effort culminated in a massive hypermedia-style XML schema for distributed and inter-connected publishing, referencing, archiving, etc. There were hundreds of elements just in the common metadata model before I even got around to defining content modules.

Then something hit me like a ton of bricks. I think it was seeing backbone.js for the first time. I realized how quickly web technology was moving forward, how easy it was becoming, and, by contrast, how absolutely shitty it was going to be to write mountains of XSLT on top of my huge DTD. Even in my attempt to create something really different, my approach was still totally clouded by the prevailing anti-wisdom of the academic publishing technology community.

I realized I was on the wrong path. It was a great thought exercise that contributed immeasurably to my day work, but I had to move on.

Happy to share more information if you like, though I'd need to dig the files out of deep storage.

I am working on something that may be of interest. Drop me a line, my email is in my profile

I know a guy. Shoot me an email, let me know what tech you know, and I'll chat with him to see if he's interested.

Reply on news.ycombinator.com