Ask HN:What do you use for searching for scientific articles

[yread]

I've been using google scholar but I'm getting tired of not getting results I want - especially the first results often seem only distantly related to my query. Plus I seem to always get an awful amount of noise - only citations or unreferencable sources.

Is there a better alternative for searching all scientific articles?

[bedris]

Papers is a great application for searching multiple article repositories and organizing your collection of journal papers.

http://mekentosj.com/papers/

[peoplerock → bedris]

"Papers" definitely looks worth a test drive (Free for 30 days; then $42/é29), though I'd love to see something web-based, not requiring download to your PC, particularly if you have to do some of your research from within institutions on their PCs/Macs. [ BTW: "papers" won an Apple award, so probably works on Macs as well as PCs. ]

[frisco → peoplerock]

It's Mac-only.

[caffeine]

To be honest, part of it is the knack of finding the right keywords. Some ideas I use:

- Hit up Google Books / Amazon for SIP (Statistically improbable phrases) related to your topic. These are useful.

- Add a bunch of +[Author name] keywords into scholar in order to get papers that cite certain people. Eventually you build up a standard list of authors that you want to find papers citing (yes, I know, this can be overly narrow, but it's effective).

- As mentioned elsewhere, use Citeseer.

- Its much easier to navigate a reference graph than to search cold. So find papers you think are excellent, and then serach everyone who cites them, all other work by referenced and citing authors, and so on.

- Use arXiv. You can set up RSS feeds for certain keywords, and search them. This is low-SNR but yields some gems.

- Most importantly: Don't search! Ask other people what's worth reading and what's good, in your group/lab/friends.

[Estragon]

What's your field? Google scholar works great for me. (Mostly genetics and bioinformatics.)

[dedalus → Estragon]

Even for CS topics I find it great as it gives me direct pdf access (I dont have ACM/IEEE account)

[simon_kun]

Basically, you need 2 things.

1) A better search interface and reference manager. If you are on OSX, Papers by Mekentosj is a fantastic option. 2) A citation search engine aggregator specific to your field. Something like Web of Science/knowledge is one of the best for my field, but I need to go via a Uni/large organisation as it's waaaay outside my price range.

Site note: In this day and age, these aggregators are a complete and utter racket. Unfortunately it's the way things are while academia finds a lingua franca other than number and status of "peer reviewed" journals for its funding orientated core business model.

[MattF]

I've always used CiteSeer for my computer vision research:

http://citeseer.ist.psu.edu/

I don't know how it rates for other fields though.

[peoplerock]

I've noticed some of same issues with http://scholar.google.com ...

It seems great if i'm trying to locate a specific article or ref by title or author ...and has in fact helped me at least get the exact citation for very obscure dissertations, conference papers, etc. But it's proved so helpful for search by keyword/topic.

[MrGunn]

I depends on your field to some degree. For life sciences, Pubmed works pretty good. You can also find some articles you didn't know you were looking for via shared bookmarks on Citeulike or recommendations via Mendeley. http://mendeley.com

[asmega]

depends on your field. eg IEEE, ACM for computer science. If you're at uni use the library resources

[Locke1689 → asmega]

Yeah definitely acm/IEEE. They also have all their publications online now and do reverse DNS to authenticate so it's all pretty easy. If I don't really have an idea of what I'm searching for then I use google. If I have a good idea but it's very specialized then I'll just walk over to the department and ask some grad student or professor where to go. If I know exactly what I'm looking for then I'll go over to the engineering library and ask someone for help.

[fburnaby]

I find it impossible to search by keyword. The trick I use is to find one relevant article via miracle, then do a reverse lookup on authors that were referenced by that paper. Once you know some names, everything becomes much more readily searchable.

[p47]

For medicine/biology http://www.ncbi.nlm.nih.gov/.

During studies I used EBSCO www.search.ebscohost.com , there is almost every field, from Astronautics to Zoology :-), but access to some resources is limited.

[craigs]

Two alternatives you could check out: http://www.scitopia.org/scitopia/ http://scirus.com/srsapp/

[sidmitra]

For arranging your personal collection of papers, you might look at Mendeley.

For searching papers, look at CiteULike

[Maro]

http://arxiv.org

Google with site:arxiv.org

[yarapavan]

www.citeulike.org - one more alternative