1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

How to categorise a list of websites?

Discussion in 'Black Hat SEO' started by jimjones44, May 21, 2012.

  1. jimjones44

    jimjones44 Registered Member

    Joined:
    May 18, 2010
    Messages:
    94
    Likes Received:
    3
    I have a database of website owners, which I know hardly any info about - not good for Black Hat.

    To enhance my database I'd like to categorise/segment my database based on industry standard categories or SIC/taxonomy codes. This will allow me to run targeted email campaigns - improved Black Hat.

    Does anybody know if software exists that can 'look' at a site's content and then suggest a suitable category?

    Thanks!
     
  2. AlexMit

    AlexMit Junior Member

    Joined:
    Jun 20, 2011
    Messages:
    122
    Likes Received:
    53
    Occupation:
    Mashups Building
    Location:
    You won't believe
    Home Page:
    How many entries in your DB?
     
  3. jimjones44

    jimjones44 Registered Member

    Joined:
    May 18, 2010
    Messages:
    94
    Likes Received:
    3
    1.2 million, so an automated system is a must. I've thought about outsourcing the job but it would take far too much time.
     
  4. lancis

    lancis Elite Member

    Joined:
    Jul 31, 2010
    Messages:
    1,632
    Likes Received:
    2,384
    Occupation:
    Entrepreneur
    Location:
    Milky Way
    Home Page:
    I doubt there is a software for that task, although it can be done using NLP techniques if you're good with programming.

    One simple technique is http://en.wikipedia.org/wiki/Latent_semantic_analysis , e.q. you count the words on each site and create a matrix X, then perform PCA (principal component analysis) - each principal component will represent a category.
     
  5. AlexMit

    AlexMit Junior Member

    Joined:
    Jun 20, 2011
    Messages:
    122
    Likes Received:
    53
    Occupation:
    Mashups Building
    Location:
    You won't believe
    Home Page:
    Idea: you could filter EMD's using niche keyword-lists.
     
  6. jimjones44

    jimjones44 Registered Member

    Joined:
    May 18, 2010
    Messages:
    94
    Likes Received:
    3
    Semantics was popping in to my head, but having no programming skills I don't think I could pursue this option.

    Thanks for the pointer though.
     
  7. jimjones44

    jimjones44 Registered Member

    Joined:
    May 18, 2010
    Messages:
    94
    Likes Received:
    3
    Please could you let me know what EMD stands for?
     
  8. assphuck

    assphuck Senior Member

    Joined:
    Feb 22, 2009
    Messages:
    1,196
    Likes Received:
    905
    EMD = Exact Match Domain

    What you are talking about doing would require a manual process for accuracy. Something to be outsourced I would say.