Hi!
I'm a German researcher currently working on ranking algorithms and automatic,
link-based detection of communities in the web. Perhaps I can contribute to nutch
in these fields. I have some ideas for algorithms and structures in mind. But
unfortunately I have neither a framework for tests nor a sufficient set of (up to date)
data. So I will take a closer look on nutch to see if I can use it for my research.
And I would like to share my results with you.
Perhaps I should give you some details about my background. I'm currently working
at the Technical University Ilmenau, Germany, in the theoretical computer sciences. I
studied mathematics (algebraic topology) and did my Dr. (Ph.D.) in this field. Right now
I'm working on definitions of communities and algorithms to detect them, and on ranking
algorithms (very recently I analyzed PageRank very detailed; the paper is in preparation).
Especially in the first field I have some ideas in mind, which may produce good clusters
of web pages representing different topics. Details can be found on my homepage
(http://eiche.theoinf.tu-ilemnau.de/~mbrinkme/en).
Regarding nutch, I especially interested in using it as a framework for experiments and
developing, analyzing and implementing its ranking and structuring algorithms (due to
my more mathematical/theoretical background especially the first two). This leads to
my main question:
Does there exist a documentation about nutch's ranking algorithm (apart from the code)? I
would like to analyze it and to suggest improvements. I think it is very important to
understand the ranking in detail to guarantee unbiased results. And, as you write on
nutch's homepage, some of your parameters are guessed. Perhaps a detailed mathematical
analysis can help here.
Yours Sincerely, Michael Brinkmeier
--
Dr.rer.nat. Michael Brinkmeier
Technical University Ilmenau, Germany Faculty of Informatics and Automation Phone: (49)3677/69-1203 Room 320, Helmholtzplatz 1 Homepage: http://eiche.theoinf.tu-ilmenau.de/~mbrinkme/
------------------------------------------------------- This SF.Net email is sponsored by: GNOME Foundation Hackers Unite! GUADEC: The world's #1 Open Source Desktop Event. GNOME Users and Developers European Conference, 28-30th June in Norway http://2004/guadec.org _______________________________________________ Nutch-developers mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/nutch-developers
