| 5:30 pm on Feb 7, 2008 (gmt 0)|
Hello Mavis, and welcome to the forums.
|The idea behind it is that there is one vote for a web page per one link to that web page. |
PageRank is a bit more complex - the "vote" of a link is weighted -- both by how much PR the linking page has and how many total links it holds.
I do understand your frustration with all the checking - but it seems to me it would be very easy to set up many Google accounts and tilt the bookmark scales.
| 5:39 pm on Feb 7, 2008 (gmt 0)|
Thanks for the welcome!
Yeah I can understand that it would be easy to skew the results - I understand pagerank is more complex, but the basic idea behind it is still that you are voting for a website.
My point is that there are a lot of non-website owning people out there and they bookmarks sites like the rest of us - and surely a bookmark is a vote for a site, otherwise, why bookmark it!
I'm just really throwing an idea around and seeing what others think!
| 6:53 pm on Feb 7, 2008 (gmt 0)|
This would be a great idea for Microsoft to employ, just slip into the Explorer user agreement that by default they are going to send bookmarking info anonymously back to home base.
I'm not sure Google would have access to a solid enough database of information to work from. There's the toolbar of course, but I'm not sure it has access to this kind of information about user behaviour.
| 7:02 pm on Feb 7, 2008 (gmt 0)|
Bookmarks are soo 20th century - these days there is no need to bookmark so long as you can find quickly site in the search engine again and again and again.
| 9:18 pm on Feb 7, 2008 (gmt 0)|
Yes, IF you remember the site exists and what its exact name is.
| 9:49 pm on Feb 7, 2008 (gmt 0)|
|so long as you can find quickly site in the search engine again and again |
Yeah, and click some poor soul's Adwords ad for the forty-eighth time!
| 9:58 pm on Feb 7, 2008 (gmt 0)|
Too easy to game the system. Plus SPAMMERS would immediately build pages to automatically stuff people's bookmarks with their sites. Plus MSIE comes preloaded with bookmarks for MSN and other MS properties, effectively giving them instant (and almot perpetual) PR of 10.
Not to mention all the related privacy issues.
| 2:10 pm on Feb 9, 2008 (gmt 0)|
I disagree with the concept completely. A webmaster (one with a good, trusted, on-topic site) would not link to a poor quality resource. Faceless bookmarking would results in the spammiest results. Getting hundreds of bookmarks is easier than getting just one authority site to link to you.
| 3:33 pm on Feb 9, 2008 (gmt 0)|
|Surely a better idea is to rank on bookmarked sites - a true vote for a website... |
If that was the scenario then I would be a 'passive voter' -- as I don't use bookmarks at all = not a single vote from me for any website (except my own of course) :)
| 7:31 pm on Feb 9, 2008 (gmt 0)|
Really? That probably sounded like a good starting premise when Page & Brin founded Google, but it's long since been proven wrong.
|A webmaster (one with a good, trusted, on-topic site) would not link to a poor quality resource. |
I don't know. The average surfer isn't likely to bookmark a site if they don't plan to return, even if you beg them to. But this does bring us around to the same problem as you have with using links as votes: Separating unnatural patterns of voting from natural, "organic" ones.
|Getting hundreds of bookmarks is easier than getting just one authority site to link to you. |
| 1:28 pm on Feb 11, 2008 (gmt 0)|
| 5:18 pm on Feb 11, 2008 (gmt 0)|
I think the basic concept of pagerank of severing up the pages that most people are likely to find by following links is sound.
Bookmark sites I have seen really never got off the ground because as LifeinAsia pointed out 'easy for spammers' However if the group of data was large enough and had enough data from regular users I would think it would make an improvement to static content. Oh, google does have bookmarks on their toolbar.
A bookmark rank that also included a trust rank would be the way to go. Since bookmarks go to static pages requiring those pages to have time in the index would discourage spammers. If it is a web 2.0 application, using the bookmarks as first results by the user for the user (or group of users) then seeing whose bookmarks get used/searched the most (as an indication of lack of spam and quality) could be a method to establish the trust/quality rank of groups of bookmarks. If a college could get involved they would likely provide considerable good quality bookmarks, I would give the professors a trustrank of 10.
| 8:32 pm on Feb 11, 2008 (gmt 0)|
|I think the basic concept of pagerank of severing up the pages that most people are likely to find by following links is sound. |
I disagree because users don't blindly follow links.
| 10:15 pm on Feb 11, 2008 (gmt 0)|
I am not saying Page Rank is best, only that it is sound in that it brings you to the pages most people have visited if they followed links. Which is a good representation of data on the internet.
I often find myself looking for things which are not of interest to the bulk of people on the internet and the page rank system works against finding that one page in a million that everybody else does not know about.
| 2:27 am on Feb 12, 2008 (gmt 0)|
Doesn't anyone here use social bookmarking tools like del.cio.us, stumblupon, digg or reddit? There are dozens of them and yes they can provide good human edited search results and yes spammers do what they can to get in but being socially based its hard for spammers to get far. its easier to fool a search engine then real people.