6 ms·
As the title suggests, the site is centered around discussions regarding scientific pre-prints and publications. Current sites supported are: * arXiv * bioRxiv
by danielecook 9y ago
As the title suggests, the site is centered around discussions regarding scientific pre-prints and publications. Current sites supported are:
* arXiv
* bioRxiv
* Pubmed
* Pubmed Central
* DOI
Threads are submitted using the publication ID, DOI, or URL of a pre-print or publication. PDFs are fetched when available and used to generate a thumbnail.
Currently, the site only supports academic email addresses. I am trying to figure out the best way to open the site to all without having to manage spam. Let me know if you have any thoughts/ideas. If you are interested in accessing the site and do not have an academic email address please PM me with an email address to whitelist.
- akavel 9y agoDid you advertise it in any academic circles, then?
- danielecook 9y agoI have mentioned sent it out to people I know and will be posting it elsewhere as well.
- JorgeGT 9y agoI'm sorry, but what is an academic email address?
- lou1306 9y agoI guess, any email address provably bestowed by a university (like .edu emails in the US).
- bringtheaction 9y agoSpeaking of which; https://github.com/leereilly/swot https://github.com/leereilly/swot Swot is a community-driven or crowdsourced library for verifying that domain names and email addresses are tied to a legitimate university of college - more specifically, an academic institution providing higher education in tertiary, quaternary or any other kind of post-secondary education in any country in the world.
- danielecook 9y agoI actually used the swot database for the site. If anyone is interested this gist will download the swot repo and generate a dictionary of the email --> school/university in python: https://gist.github.com/danielecook/19793d2faad66d83cef2514c9760f00c https://gist.github.com/danielecook/19793d2faad66d83cef2514c...
- JD557 9y agoUsually when you study/work on an university, you get an email address with the university domain (e.g. jorge@mit.edu). Those are used in many places as a simple way to validate that you are a student or a researcher.
- avian 9y agoThe important part here is "usually". It should be stressed that using lists (such as SWOT mentioned in another comment here) can add to the bias towards US and other well-known universities. I have been involved in academia for 10 years now, and I never held an email address that would match against the SWOT list.
- JorgeGT 9y agoOps, I meant to ask how they were validating academic addresses. I'm actually a postdoc, but my university domain is @upv.es, and I've found that sometimes it doesn't get recognized in American websites which expect a .edu for claiming academic discounts or things like that. @bringtheaction comment mentioning SWOT is an interesting solution. Although the name of my institution is outdated! I'll send a PR.
- afandian 9y agoThis looks really cool. I've emailed admin@upvote.pub about tracking your use of DOIs.
- kiliantics 9y agoGreat job! I had thought of building something like this before. Some ideas I'd toyed with were: - A tool for annotation so that people could comment on specific portions of papers. This could help new researchers in a field to more quickly understand the different broader contexts in which a result can be interpreted, or the technical limitations of a given method and how that should be considered when say using the results to motivate other work. - An option for some form of anonymised commentary, so that people could voice their views on a particular work without fear of reprisal. I think this would be amazing for non-tenured academics to exert some form of power in the community, if a platform like this were to take off. Of course, it would make for potentially very messy and difficult to moderate discussions.
- throwaway080383 9y agoThe first thing you mentioned, annotations, or maybe even inline Q&A/discussions, was my dream in grad school (math). I imagined something like a separate MathOverflow for each paper. The basic problem is that I was constantly finding myself struggling with a tricky part of some paper from the 70's ("Why is this sentence true??"), and I knew many students before me had struggled with the same section, but there was no way for one really great expositor to clarify it, get upvoted, and be done.
- claudius 9y agoswot seems to be restricted to places of higher education, leaving out research institutes, such as the MPG association in Germany or the national labs in the US.
- o2348diuu 9y agoI completely understand the issue of not wanting spam of various sorts broadly defined, but I do think restricting it to academic addresses is a huge mistake. As an established researcher on the verge of leaving an academic institution, but who is still very much interested in research, I'm becoming acutely aware of how cliquey academics is, and how problematic it is for the field and society as a whole. The assumption that anyone who might have anything important to say would be at an academic institution is extremely dangerous (just as is the problem of access to journals outside of academic institutions, etc.).
- afandian 9y agoSeconded. I'm not an academic but I deal with academic publishing and I might want to participate. My suggestion: use ORCiD.org for authentication with OAuth. This is a researcher profile. Most academics have one (or should!). They're free, the service is open. I think it strikes the right balance between barrier-to-entry for spam and accessibility.
- icebraining 9y agoHow does ORCiD.org itself deal with spam? Seems like they just use a captcha, no?
- killjoywashere 9y agoMore about ORCiD for the skeptical: https://en.wikipedia.org/wiki/ORCID https://en.wikipedia.org/wiki/ORCID
- afandian 9y agoIf anyone has any questions I can ask someone from ORCiD to come over here and answer them. They're very open!
- Vinnl 9y agoI've implemented sign-in using ORCID, and although I support the idea, its UX is far from great, its documentation isn't either, and it also isn't terribly stable (it was down on my launch day...).
- killjoywashere 9y agoCan you please allow .mil and .gov addresses? Those "government scientists" add up to a lot of researchers.
- euid 9y agoTotally agree. I did government research for a while and published papers. Don't exclude them - anyone at the intersection of government work and research is likely to have meaningful contributions to this website.
- khr 9y agoDo you plan to add PsyArXiv? (https://psyarxiv.com/ https://psyarxiv.com/)
- throwawayjava 9y ago> Let me know if you have any thoughts/ideas Many papers list authors' email addresses. Scrape PDFs on the services you support for email addresses. If a domain name occurs in enough author email addresses and isn't on a "known generic" list (gmail et al), consider it "safe".
- lars 9y agoOnly 39% of the submissions to the latest ICLR (machine learning conference) were from .edu emails [1]. There's a lot of research going on outside of academia, and there's also a large portion of academia that doesn't use the .edu TLD. That said, I don't have a solution to your problem of fighting spam. [1]: http://webia.lip6.fr/~pajot/dataviz.html http://webia.lip6.fr/~pajot/dataviz.html
- semi-extrinsic 9y agoHow about doing a scraping of e.g. 10k Google Scholar profiles with more than ~10 publications to get the domains of their verified email addresses? Then you have a much broader whitelist than just *.edu. Or you can crowdsource this, say I want to register as john.doe@instituteoptique.fr; in the registration form simply require me to give a link to someone's Google Scholar profile with an email address from the same domain. It should also be noted that e.g. on Figshare, anyone can upload any PDF and get a DOI for it.
- liamcardenas 9y agoHave you actually had problems with spam? Or is this a purely hypothetical problem?