Joho the Blog » Google Autolinker, part 2
EverydayChaos
Everyday Chaos
Too Big to Know
Too Big to Know
Cluetrain 10th Anniversary edition
Cluetrain 10th Anniversary
Everything Is Miscellaneous
Everything Is Miscellaneous
Small Pieces cover
Small Pieces Loosely Joined
Cluetrain cover
Cluetrain Manifesto
My face
Speaker info
Who am I? (Blog Disclosure Form) Copy this link as RSS address Atom Feed

Google Autolinker, part 2

In case you were wondering, this is how Philipp’s autolinker autolinks my post about his autolinker:


Philipp Lenssen’s Google Blogoscoped shows the results of a tool that checks Google for the longest occurrences of strings from a text on the web, and then turns it into a link. I. e. , if you run your page through it, it finds any phrases that have a significant number of hits in Google, and links each phrase to its top return. (I think. ) Cool idea. Alas, Philipp writes that the tool is too slow to be made public, although he offers to run it by hand if you send him email, which I hope for his sake vast numbers of people don’t. [Thanks, Hanan, for the link.]


[Technorati tag:]

Previous: « || Next: »

3 Responses to “Google Autolinker, part 2”

  1. I think that is very clever and all; but dont you feel that it places just 1 to many links on your post?

  2. Oh and I forgot 1 more point I wanted to make. You do not know what will be created into a link, and what you will be linking to. For me, when I am writing a post I like to plan out my links, directing readers -not like I have a lot- to were I want them to go. I think the randomness of the links would get on my nerves after about 2 entries.

  3. Brad, if I can jump into here: my tool was meant not so much as “this is how people should link in the future” but as a way to get you thinking in the direction of certain new tools and/ or ways to analyze text. I wrote an update in my blog where I put it to rather good (experimental) use:
    http://blog.outer-court.com/archive/2005-02-04-n35.html

    I do often find a “googlonym” of specific text would have fared better, e.g. when I would write “Yahoo buys Fred Miller’s Company XYZ” then there’s be no inherent need for me to link either “Yahoo”, “Fred Miller” or “Company XYZ” because all those things could be googled by the reader (or a tool) anyway — also, a “googlonym” (also called a memomark) fares better over time because the old URL might be gone after a while.

    Of course there’s a somewhat paradoxical situation here. If everyone would use links like this, it would be a feedback loop: search engines could not really decide anymore what’s relevant, and existing sites would never leave their top ranking.

    In a way, this is already happening today: when you want to link to an Einstein biography, any biography (you are just looking for a quick link) you may google for “Einstein Biopgraphy”, and then just use this page as link target… thereby increasing its strength. That doesn’t mean the page is the best biography on good old Albert. Maybe the best biography was this terrific piece on result page 12.

    So sometimes, it helps if indeed we take good care to place the best manual links we can…

Leave a Reply

Comments (RSS).  RSS icon