finding cutting edge papers and books

Hi all,

what are the best strategies to find cutting edge papers and books on a field of mathematics?

..

Example:
2-3 years ago I had to analyze a time series. I found a paper and showed that to a mathematician who referred me to the REAL state-of-the art method how to do it.

Then I read the book ‘Analysis and Probability: Wavelets, Signals, Fractals’ by Palle E. T. Jorgensen, which is excellent and a good reference.

How can I find for example the latest advances in (applied) mathematical Logic (modal logic, high order logic, type theory, proof theory)?

How can I find the state of the art advances in time series analysis since mid 2007?

My current strategies are using google search and google scholar to find papers. Or to try to find a good mathematician in a field an ask in person for papers.

Thanks for any Tipps!
Steve

-
Regarding Logic: are categorical logic and type theory/dependent types state-of-the art topics in logic at the moment? (I'm new to logic..) – stmath Oct 22 '10 at 15:49
Some of those fields (proof theory, for instance) are fairly small, so your best bet may be looking at the web pages of some of the people working it; in addition to finding their latest papers, you can probably find some survey papers outlining what's been going on in the field. – Henry Towsner Oct 22 '10 at 17:33
stmath: yes, they certainly are. In some places, at least. – Adam Oct 23 '10 at 1:01
Regarding your original question, you really should to narrow it down to one field; every field has its own journals and conferences. May I suggest that you re-post just "How can I find the state of the art advances in time series analysis since mid 2007?" by itself as a question? I bet you would get some helpful answers if you kept it specific like that. – Adam Oct 23 '10 at 1:02
I'm wiki hammering this question since there is no single correct answer. – François G. Dorais Oct 24 '10 at 17:17

One method that is maybe more indirect is to regularly browse through the new articles posted on the arxiv in the section of your interests. I use an RSS reader for this, but you can also subscribe to an e-mail list I think.

Also, conferences for example are always a good way to keep up to date of what's going on in your field.

-
Hrm, I have to say, though... it's probably going to be hard for a newcomer to sort out the good from the bad on the arxiv (it's just an archival service after all, not a peer-review mechanism). I've found that it's not very useful for me to follow the arxiv RSS feeds except in one or two very narrow areas where I'm confident I can tell the nonsense from the insightful... for the many more topics I'm casually interested in I haven't found the RSS feeds all that helpful. – Adam Oct 23 '10 at 1:05
it sounds like a good approach and a alternative to what I am doing now. Sometimes I just wonder: science people are supposed to be smart. why are ALL the interfaces to sites like arxiv so poorly designed? I am searching also on 'Zentralblatt MATH' and again the interface is horrible.. – stmath Oct 23 '10 at 7:57
@Adam: yes, you are right. This mainly works if you already have some familiarity with the field. This is more an answer to the "cutting edge" psrt of te question, not the best way to learn anew field. – Pieter Naaijkens Oct 23 '10 at 8:33
stmath: that's funny, I really love the arxiv interface; it's so simple and everything is quick and snappy and responsive. I'm happy they haven't junked it up with all sorts of unnecessary flashy javascript nonsense... – Adam Oct 23 '10 at 19:48
I also love simple interfaces, hate flash and most javascript. However designing the page, for instance with reference to the 'less and more' principle (by dieter rams), is a must for a quality website. – stmath Oct 25 '10 at 7:27

When I start getting acknowledged with a new field, I typically use several search engines (general purpose or scientifically oriented) to find publications that seem relevant, then go forwards and backwards in time (that is, I look at cited papers and citing papers) to try and find surveys of the field. This allows me both to retrieve latest papers and to identify relevant ones as most frequently cited.

-

The best way to find cutting edge papers is to attend conferences and see talks about current research. Research journals lag years behind current research. Even preprints often lag behind announced results - in some fields, including logic, people often speak about results that are not yet available as preprints.

In logic, the arxiv is not extremely useful for finding new results, because only a minority of practicing logicians publish preprints on it.

Logic has very little to do with time series, however, so the conventions may be different for research in time series.

-

There are some good resources mentioned already on this site. In particular see this question: Most helpful math resources on the web

Ryan Budney makes a comment under Thanos D. Papaïoannou's response about MathSciNet. Given any relevant paper on the topic of interest, you can find papers that cite it.

This question may also be worth looking at: Free, high quality mathematical writing online?

-

Web links at UCLA and NIST (USA's National Institute of Standards and Technology):

http://ces.stat.ucla.edu/software/time-series-analysis

Introduction to Time series Analysis (Engineering Statistics Handbook) at http://www.itl.nist.gov/div898/handbook/pmc/section4/pmc4.htm

It's probably best to use a multi-step process. Search engines can point you in the direction of papers with keywords or title words which you may be interested in. Then use those papers and look at the bibliographies of those papers.

The bibliographics references will lead you to a set of authors who publish in that field. Now search for further articles by these authors, and for their academic or work-related web-sites. Their own bibliographies on their web page may help you to find other papers and other researchers working in that field, and also lead you to conference presentations which they may have done.

The most "cutting-edge" work is going to be the work recently presented at conferences, working groups, and SIGs (special interest groups, such as the subgroups of IEEE, e.g. SIGGRAPH: special interest group in graphics, which has a yearly convention in Los Angeles). Once you've found a listing of conferences, you can look for the collection of conference proceedings or talk title, and figure out the research groups and individuals involved in working in the topics of your interest.

I hope this helps you.

-