Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Why not imagine it in a way that works?

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



"Learning Search Filter"

Learning Search Perpetual Filter
  (+2, -1)
(+2, -1)
  [vote for,

This program would be designed to tag and store "learn" the url of a web page that has been viewed and as an option "keep" the page from returning... in search results... ever again. (unless disabled)

I have been searching for so many things for the longest time and I wish I could filter out the pages I have looked thru... ignore them "automatically" and maybe find what is "new" and maybe what I am actually looking for...

The search "advanced preferences" features in the top 10 search engines just dont cut it...

sound like a good idea...?

oxygon, Dec 05 2004

http://scholar.google.com/ New google search engine for academic / technical papers. Surprisingly good, I thought. [jutta, Dec 05 2004]


       Perhaps if it also ranked sites by accuracy and not how many sites linked to an url (any with an *.edu get plus 1, any well known like bartlett's, wikipedia, halfbakery +10) with customizable search groups (search for "War" in Fact, Humor, News, History, Science....)
JeremiahBritt, Dec 05 2004

       Two problems.   

       1) It's tough to make users do something for each result. If you read a lot, you're probably fast; and just having to toggle a checkbox for the things you want/don't want gets annoying, with a fairly high penality for false deletion (you stop finding the thing you knew was there) and a high incidence of justified deletion (most of the results are crap).   

       2) If you get a lot of junk, there may be other, faster things you can do to improve your search result quality in general. I don't mean knowing the latest keyword: prefixes; more things like picking up keywords from the contexts you mean, and suppressing some from contexts you don't mean, and narrowing things down conceptually that way.   

       That said, yes, absolutely, "I have seen this page before" or "I've been to this site a lot" would make for a great search criterion - both for inclusion and for exclusion.
jutta, Dec 05 2004

       Seems to me you'd accomplish what you're suggesting with a [-] library, that would act a "standing filter" for your ongoing searches. You'd have to decide at some point that you'll never have use of, say, "muscle spasms" as a reference statement in current or planned reference work and you would add "muscle spasms" to the [-] library. You would thereby evolve a list of words and phrases of the form -"muscle spasms" that your search engine would call as a ubiquitous filter.
reensure, Dec 05 2004


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle