Hi ajg I think for the best reliability, another tool would be better - something that gives some sort of ranking of the similarity of two phrases. No idea what tools are out there though.
Sorry I can't be of more help. -- Pat On 10/12/2008, at 12:23 AM, agib wrote: > > Hi, and thank you for the responses! > > What I mean is this: > > So instead of posts, think questions. Imagine a site like Yahoo! > Answers (http://answers.yahoo.com/). Near the ask box, there's a > statement: "Can't find it with search? Ask" > > I find this flow for the user a little cumbersome, and it relies on > the user not to be lazy and search before asking. I was thinking > wouldn't it be nice if the search step was just part of the asking: > that is, a user asks a question and the app searches for similar > questions and says "are you sure your question has not already been > asked?" The point being that the question doesn't have to be > identical, but rather, if it is similar to other ones it encourages > people not to ask questions /similar/ to those that have already been > asked. > > For now I'm thinking of :match_mode => :any. However, it seems like > this isn't necessarily the best solution for the situation... ideally, > the order of relevance of the search results would be weighted by the > *least* common words in the question. Anyway Sphinx might not be the > right solution. A separate word count table in the DB might be > necessary to weight search results. > > Anyone seen anything like this? > > Thanks again! > > -ajg- > > On Dec 9, 4:56 am, "Henrik Nyh" <[email protected]> wrote: >> If you really do need to use Sphinx for this, perhaps phrase search >> (quotes,http://sphinxsearch.com/docs/current.html#extended-syntax) is >> what you want. >> >> It's hard to know what you're looking for without more details. Do >> you >> want to avoid post bodies being exactly identical? If so, >> validates_uniqueness_of or something similar that doesn't use Sphinx >> makes more sense. If you want some more refined similarity measure, >> let us know what the requirements are exactly. >> >> On Tue, Dec 9, 2008 at 9:36 AM, Henrik Nyh <[email protected]> wrote: >>> class Post < ActiveRecord::Base >>> validates_uniqueness_of :body >>> end >> >>> ? >> >>> On Tue, Dec 9, 2008 at 7:10 AM, agib <[email protected]> >>> wrote: >> >>>> On my site I want to prevent people from posting the same thing >>>> twice... so I want to run a search once the user is done with the >>>> post >>>> to prompt them to make sure their post is unique. Using the entire >>>> post as a search query doesn't work... should I break it up word by >>>> word? I'm using sphinx with thinking-sphinx... has anyone >>>> encountered >>>> this before? >> >>>> I was thinking of searching all the words individually and picking >>>> merging the smallest search result sets... >> >>>> Thanks, >>>> Aaron > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Thinking Sphinx" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/thinking-sphinx?hl=en -~----------~----~----~----~------~----~------~--~---
