Jeremy Liew of Light Speed Ventures has been writing a series of posts on meaning and structured information. Not only these posts are very well researched, but they are also really well written. Jeremy is skilled at getting to the heart of the issue and at translating from a geek speak to the end user value.
His original post, introduces the problem and lays out three approaches that transform unstructured information into structured:
1. User generated structure
2. Inferring structure from knowledge of the domain
3. Inferring structure from user behavior.
The follow on post discusses the user generated structure. Jeremy talks broadly about tagging, asking users for a structured input, bottom-up semantic technologies like RDF and OWL and finally generating a structure via central repository of meaning.
The next post, which he just published, focuses on inferring the structure from the domain knowledge. This is something very dear to me personally, because this is the top-down approach that we’ve been working on, here at AdaptiveBlue. Jeremy has an excellent explanation of how applications can take advantage of a vertical.
The combination of a focus and understanding of the semantics allows services like vertical search engines, specific social networks and our own smart browsing technologies excel at weaving structure on top of unstructured information. My recent post about SmartLinks seconds what Jeremy is saying and in addition shows that the top-down approach can be scaled to the whole web.
I am following this series and actively commenting on all posts. I encourage you to do the same. With more posts to come, this series is already shaping up to be an important meditation and a forum on the subject of structured web.