Dear Wiki user, You have subscribed to a wiki page or wiki category on "Nutch Wiki" for change notification.
The "NewScoringIndexingExample" page has been changed by LewisJohnMcgibbney: https://wiki.apache.org/nutch/NewScoringIndexingExample?action=diff&rev1=6&rev2=7 '''N.B.''' This page and the functionality described within is only applicable and relevant to Nutch 1.X. + <<TableOfContents(4)>> + + = Introduction = + Below is an example of running the new scoring and indexing systems from start to finish. This was done with a sample of 1000 urls and I ran two different fetch cycles. The first being 1000 urls and the second being the top 2000 urls. The loops job is optional but included for completeness. In production we have actually removed that job. This was done with a clean pull from Nutch trunk as of 2009-03-06 (right before 1.0 is set to be released). If anybody has any problems running these commands or has questions send me an email or send one to the nutch users or dev list and I will reply. Please send it to kubes at the apache address dot org. + = Workflow = {{{ bin/nutch inject crawl/crawldb crawl/urls/ @@ -18, +23 @@ }}} One thing to point out here is that WebGraph is meant to be used on larger web crawls to create web graphs. By default it ignores outlinks to pages in the same domain, including subdomains, and pages with the same hostname. It also limits to one outlink per page to links in the same page or the same domain. All of these options are changeable through the following configuration options: + + = Configuration = {{{ @@ -47, +54 @@ </property> }}} + + = Additional WebGraph Classes = But by default if you are only crawling pages within a domain or within a set of subdomains, all outlinks will be ignored and you will come up with an empty webgraph. This in turn will throw an error while processing through the LinkRank job. The flip side is by NOT ignoring links to the same domain/host and by not limiting those links, the webgraph becomes much, much more dense and hence there is a lot more links to process which probably won't affect relevancy as much. @@ -163, +172 @@ bin/nutch org.apache.nutch.indexer.field.FieldIndexer -fields crawl/fields/basicfields/ -fields crawl/fields/anchorfields/ -output crawl/indexes }}} + = Class Diagram = +
