Could you please explain the difference of this planned module to the already existing Content Recommendation Engine?

Comments

marcor’s picture

marcor commented

Please explain Content Recommendation Engine, then I explain the differences! ;)

Ok I read the module description and the algorithm, and as far as I understood, CRE is used to rate content on an absolute scale, so a high value means: "Whoever you are and whatever you believe, this rated content is good. But I don't tell you, if it might be interesting for you." High values come from votings on a given cardinal scale ("1,2,3,4,5...") to express the value for you, so that it is estimated what value it wil have to anybody.

Neighbours module doesn't rate persons or things. You don't rate with numbers, you only use tags that are values on a nominal scale. You don't rate "good" or "bad", you just describe. If someone becomes your neighbour, then this depends only on you and her and a high value on her for you means: "To me, this person is similar. To you, she might be completely different. And knowbody knows if this is good or bad."

Note: This module is not written yet, but I will start soon. Any help is very appreciated. If you think, I'm wrong and CRE might also cover my objective, please tell me!!! ;)

marcor’s picture

marcor commented
Assigned: Unassigned » marcor
Priority: Normal » Minor
Status: Active » Closed (fixed)

Just thought over what I wrote. I think the main difference is: CRE depends on votings, Neighbours on tags.

For this module is not written, yet, the documentation is targetted at developers. I like to keep it lean until it is released.

More suggestions are welcome, of course.