[p2p-hackers] Getting started

[Wolfgang Mueller]

Dear all,

I am a P2P similarity search person, and have some other literature
suggestions (as suggestions always are, the're incomplete):

> Btw... Kademlia, Chord and Pastry are all structured by design, so you might 
> want to read a paper on unstructured networks too, where topics like random 
> searches are covered. I dont know of any good paper, but maybe someone else 
> does?

In terms of unstructured networks, I am a great fan of this stuff 
http://arxiv.org/abs/cond-mat/0406152 . It is data type agnostic, yet
efficient, surprisingly efficient, if you consider that the
preconditions to make it work are very weak. 

In addition to Neurogrid, there are plenty of other approaches out
there that try to improve link structure based on semantics. The
Firework Query model springs to mind:
http://citeseer.ist.psu.edu/ng02peer.html and its refinements done in
Bibster http://bibster.semanticweb.org/publications/publications.htm ,
http://citeseer.ist.psu.edu/context/2394099/715233 however, I guess
some of Bibster's authors are on this list, right?

In terms of DHT literature:
http://www.srhea.net/papers/bamboo-usenix.pdf I like very much, as it
provides great level of implementation detail. 

For P2P-IR the great classics IMHO are the papers about psearch and
successors http://citeseer.ist.psu.edu/tang02psearch.html . 
A general view on similarity search using one approach given there is
given in: http://pdos.csail.mit.edu/~rtm/papers/search_feasibility.ps .
One approach to solving this problem is PIER:
http://pier.cs.berkeley.edu/papers.html .

A quite different approach to P2P-IR is PlanetP
http://www.panic-lab.rutgers.edu/Research/planetp/
with scalable extensions Minerva
http://www.mpi-sb.mpg.de/departments/d5/software/minerva/
and Rumorama (our pet here in Bamberg)
http://portal.acm.org/citation.cfm?id=1099706&coll=ACM&dl=ACM&CFID=6

And I won't forget mentioning the Freenet-based FASD by Amr Kronfol. 
What is the current state of this, by the way?
http://citeseer.ist.psu.edu/571354.html

Disclaimer: I know this is pretty incomplete and rather cited at random, 
I am aware of plenty more papers e.g. at Stanford Peers and other "usual
suspects ". Please view this rather as a starting point.

Cheers,
Wolfgang

--
Dr. Wolfgang Mueller
LS Medieninformatik
Universitaet Bamberg