The "Similar Content" Section
Some knols may show a "Similar Content on the Web" section, which may show a link to another site that has content 50% to 100% similar to what is on a knol.The Similar Content section is not continuously updated. It may be refreshed once every week or two.
The "Related Knols" Section
Some knols will automatically show up to five "Related Knols", below the "Also Wrote" list. This is intended as an aid to browsing and discovering knols about similar content.Related Knols are chosen by an algorithm that looks for pages with similar titles or content, written by different authors. There are many factors which may go into the selection of Related Knols, and the algorithm is subject to change at any time.
This list is populated automatically. The presence of a link within the Related Knols section does not mean that the author of the knol has in any way endorsed or agrees with the content linked to.
The Related Knols pages may update every few days or weeks, as the content, popularity or other attributes of pages change over time.
Removing knols from Related Knols
Authors can remove links to pages they find objectionable, offensive -- or just simply unrelated. Click on the "Remove this link" shown below the related knol and confirm that you want the link removed.You can turn off the Related Knols section, as described here.
The "Also Wrote" Section
Once you have more than one knol published, all of your knols will contain an Also Wrote section, where we will show links to up to five of your knols. These links help readers discover what you have written, and may help search engines find all of your pages.You can turn off the Also Wrote section, as described here.
Peter Baskerville
Invite as author
Different metric needed
For this reason I would prefer that the metric be changed slightly to show "Similar Content when Published". In other words, when this Knol was published, the following other sites already had similar content.
At the moment, copies or our Knols to other sites are getting mixed up with copied content from other sites. I vote for just making the metric report on inbound content, not on outbound. Otherwise we may inadvertently tarnish the reputation of a quality Knol author who has had their work copied to another site and being seen as a plagiarizer (when the reverse is actually the case)
You are correct that at the moment the "Similar content" links just mean "Similar content", they in no way imply directionality of copying. It is left to a reader's discretion to interpret which came first, the chicken or the egg.
To make a stronger claim, X copied Y, that would actually be quite difficult to make with a sufficient degree of accuracy, for a variety of technical and/or policy reason. We do not expect to be doing that any time soon.
EditSaveCancelDeleteDeleteBlock this userReport abusive commentHide report window
Norman Creaney
Invite as author
Similar knols
EditSaveCancelDeleteDeleteBlock this userReport abusive commentHide report window
What we will be doing in the short term, is dropping the percentage from the "similar content" display. We will still show the link, but without the 90-100% numbers it should have more of an informative rather than accusatory connotation.
The problem is -- that number is something closer to being "the probability that these two pages share some notable amount of unique language in common" -- it is not the stronger claim people have been understanding it as, "the proportion of content these pages share in common.")
EditSaveCancelDeleteDeleteBlock this userReport abusive commentHide report window
I think it is a good thing that you try to identify where similarities exists - I'm certainly against spamming, etc. I just think that the numbers were out of place on my knols, in which the similarities were perfectly innocent & justified.
What you now suggest sounds like a fair compromise.
EditSaveCancelDeleteDeleteBlock this userReport abusive commentHide report window
Knol Help
"Similar Content" is mistaken if you have a page that redirects to a knol.