The End of Google?

OK, that's a little bold, but there is a new trend of organizing information on the web that I feel will ultimately trump Google's algorithm. The trend is called Social Tagging or Folksonomy. Essentially it involves asking users to "tag" (i.e. bookmark) various web pages and assign key words to them. Instead of the bookmarks being stored in their browser's favorites menu, they are stored online. Sites like del.icio.us and ma.gnolia.com have made this trend wildly popular.

This system offers several advantages over traditional bookmarking. First, since your bookmarks are stored online, you can access them from any computer (i.e. they aren't tied to your browser. Second, bookmarks/tags can be shared with others. So you can see your friends' bookmarks and they can see yours. Most importantly, tags can be aggregated together and plugged into a search engine. This system has huge advantages over traditional search engine spiders (i.e. Google).

Tag-based classification is done by human beings, who truly understand the content of a web page, as opposed to spiders which algorithmically attempt to determine it's meaning. The more people who tag a web page with a particular word, the higher the tag-based search engine will rank it. This provides for semantically classified information, which is much more dependable and powerful than a computer's interpretation of what content means.

This model can easily be brought into the Enterprise with very little investment. A simple distributed bookmarking service can be created that allows your employees and/or clients to tag any content that exists on your internet/intranet. This information can be aggregated together to be fed into your enterprise search engine, or a new tag searching service can be created. There's no reason to retire your search engine spidering on day one, but over time as more and more people see value in using the tagging service to share tags with one another and accessing their tags from any computer, that aggregated categorization will trump your search engine's spidering interpretation of your firm's proprietary content. It also allows your employees to you use your firm's vocabulary for classifying information (i.e. think acronyms). I mean, what algorithm is going to know that a particular document is the EAM to support the TAS product's RUP requirements for the Vertigo release?

-J