The obvious algorithm is:- For this tag (tag1) find all the tags (tag2) from this vocabulary that are used on the same nodes. Then create term_relation entries for tag1, tag2 and tag2, tag1
And that'll work for a totally automated method, yeah, fine with that. For my particular needs, however, the "similar keywords" code is not one of automatic deduction, but one of usage reduction. The (currently implemented in a homebrew) system I need (have) is visualized here: http://disobey.com/detergent/2005/similar_keywords.jpg Since I can't trust (my) people to understand the vagaries of what a "synonym" is, I'll be saving information into the "related" field. -- Morbus Iff ( you are nothing without your robot car, NOTHING! ) Culture: http://www.disobey.com/ and http://www.gamegrene.com/ Spidering Hacks: http://amazon.com/exec/obidos/ASIN/0596005776/disobeycom icq: 2927491 / aim: akaMorbus / yahoo: morbus_iff / jabber.org: morbus