3 ms·
Didn't the article say the vetting was mostly academics doing it for free? Shouldn't cost much then. Curating can be a paid service as was the case in the artic
by nickpsecurity 10y ago
Didn't the article say the vetting was mostly academics doing it for free? Shouldn't cost much then. Curating can be a paid service as was the case in the article. Whereas, I know what the cost is: SciHub is $5,000/mo to operate if all one does is provide a collection of papers. Distribution to a world-wide audience would cost more but could be absorbed by universities mirroring things and using storage appliances like that OSS one that shows up here periodically. Not to mention cheap boxes that cache the hot docs in RAM.
People that want fancy shit can pay for it. Meanwhile, I got 16,000 papers on IT and INFOSEC out of ACM and IEEE through my university just using a search box with titles, authors, abstracts, and PDF download button. All I need to bring cutting-edge and prior good work to Hacker News regularly. Well, summaries of it unless I find a public download that allows me to share. Esp CiteSeerX, a perfect example of what I'm proposing. It's the bomb. :)
- IanCal 10y agoPeer review is done for free yes, but you also need to manage that process and there's a filtering step before you send things to peer review. Then there are checks like someone going through the references and making sure they're all there, that dois are assigned and assigned correctly, etc. There's the software for managing all this process too, and then all the extra cost around hiring people like management, HR, office space, etc. You can look at all of these and say "yes but that's not hard it's just..." but the fact is someone does work and that'll need to be paid for, so there is some cost. Often only papers that are accepted actually pay the cost, so they're effectively paying for all the work done on papers that don't make the final cut. Nature, if I remember the figures right, accept about 5% of submissions. So for every paper, there's a certain amount of work that needs to be done on 19 others. I'm not justifying any particular cost, but it'd be worth looking at PLOS and their fees as they're a non-profit. > SciHub is $5,000/mo to operate if all one does is provide a collection of papers. Is it reasonable to put the cost of running a publication at near or under one FTE? You'd at least have to do more if you had search and weren't just piggybacking on google scholar.
- nickpsecurity 10y ago"Is it reasonable to put the cost of running a publication at near or under one FTE? You'd at least have to do more if you had search and weren't just piggybacking on google scholar." Where did you find they're just piggybacking on Google Scholar? I know nothing of the implementation outside what she spends for a specific amount of papers. "Is it reasonable to put the cost of running a publication at near or under one FTE? You'd at least have to do more if you had search and weren't just piggybacking on google scholar." Oh yeah, I agree. A survey of many looking at various attributes with associated costs is probably in order to help establish a baseline when predicting costs of larger activities.
- IanCal 10y agoAh search appears to be down, and possibly since Feb. Most news articles talked about them using google scholar and I'm sure I remember using it and it taking me to a google scholar page but with altered hyperlinks. If you search around you'll find other people describing something similar. I assumed it was still like that but apparently google have blocked them [0] > Oh yeah, I agree. A survey of many looking at various attributes with associated costs is probably in order to help establish a baseline when predicting costs of larger activities. Yeah it'd be very interesting to know how much something like this might cost. PubMed might be a good analogy, if there was an EU or worldwide version run like that which had proper backups, etc that'd be very interesting. [0] http://bigthink.com/neurobonkers/a-pirate-bay-for-science http://bigthink.com/neurobonkers/a-pirate-bay-for-science
- dougmccune 10y agoAs another commenter alluded to, the cost to host PDFs is barely a factor in the cost of the whole process. The bulk of the cost is the paid staff and the systems required to run the submission and peer review process. Yes, reviewers review for free. But often there are paid editors and there is always significant staff to handle the submissions, find appropriate reviewers, follow up on the process, and ensure that the whole thing actually gets done, check for plagiarism and fraud, etc. That's not to say it's a particularly good system. The whole process of submission, revise and resubmit, rejection, submit to another journal, etc can take many many months (or even years) and traditional peer review isn't perfect (or even close to it) by any means. But arguing that there are no costs involved is simply false. If we look at non-profit publishers like PLoS we see processing charges between $1,500 - $3,000 per article. And you have to realize that there are multiple kinds of peer review, at a basic level it breaks down to two categories, what I'll call "methodological review" and "impact review". PLoS One, for example, only does methodological review. So at no point in that process is anyone trying to assess the importance of the paper. eLife is a newer open access publisher going a bit the other way, trying to have a journal selecting for importance of the caliber of Nature, Cell, etc. Their initial costs (these numbers are a few years old) came out to be about $14,000 per article. Depending on how you approach peer review and what services you are trying to provide you can certainly get the cost down, but I have yet to see a model that can get down below about $700 per article (still only doing methodological review). Sci-Hub can be cheap because it piggy backs on the process after the vetting and curation are done. But that's like comparing apple to oranges.
- nickpsecurity 10y ago"Sci-Hub can be cheap because it piggy backs on the process after the vetting and curation are done. But that's like comparing apple to oranges." Thats actually the most important counter in your post. It's a good point as they do effectively piggyback on the curation process. I think there's still a smaller comparison to be made if we look at the web. All kinds of people post all kinds of stuff. Good stuff eventually makes it into paid or free places that curate it for various communities. Yet, I can access any of these public sites at any time. I can look myself, follow a recommendation made on a forum, go to a free site offering links to good resources, or go to paid site claiming higher quality. Whereas, with main publishers in academia, I have no such options: paid, expensive, and for a long time. Let's get to the costs you describe, though, as it's the meat of your counter. "If we look at non-profit publishers like PLoS" Doing that was going to be my proposal as they filter out the fat that comes with management-heavy for-profits. Good you beat me to it. ""methodological review" and "impact review". " Do you have a link to what goes into methodological review, esp good review, so I can try to assess how to reduce its cost? "processing charges between $1,500 - $3,000 per article" That should be easy to cover with non-exhorbitant, subscription models plus some charge to submitting organization. Such articles, after costs recovered, should become free with journal subscription. I think your organization is already doing something like that with one of the startups, though. "came out to be about $14,000 per article." This is where publishers won't be able to absorb the costs unless they're charging an arm and a leg. I think the article's recommendation of it coming out of grant money of submitting organization is best one. Want to get prestige? Then review is a cost of doing it. Could be uphill battle but I agree nobody can offer anything free unless such costs are covered externally or with charges for a period of time.