I came across a good microformats online presentation by Kevin Lawer (highly recommended). For me the idea that really resonated was “Microformats can be the semantic web today, with existing user behavior and using existing browsers and tools”. I agree with that – IMHO, a Semantic Web can be built only from the bottom up, as most large scale Web innovations come from simple tools or standards which don’t demand too much effort from content creators or cosumers to use.
Since the Semantic Web seems to me to really be about better search results, the next obvious step seems to intergrate Microformats with search, thus making it possible to search for, say, reviews for Mexican restaurants in New York. Unless I’m missing something, this doesn’t really exist yet, except in specialized sites. Microformats markup is contained entirely in the XHTML tags, which keeps it clean and elegant but also hides it from most search engines’ crawlers. The data itself isn’t hidden, obviously, but the fact that it is formatted is.
One way to fix this might be to add some search engine-visible text to pages containing microformatted data. Perhaps a small blurb in the footer, like “Microformat hReview” or similar. It’s possible to use a
display: none to make the text invisible to the user yet indexed by search engines. It’s obviosuly not as clean as putting everything in the XHTML, but it allows discovering the content via Google or other major search engines by looking for “hReview Mexican New York rating 5”, for example.
Now, clearly this isn’t an end in itself. I don’t expect many people will be including “hReview” or “hCalendar” in their Google searches. But this will allow sites that will aggregate the microformatted data via Yahoo or other search engines (now that the Google Search API is deprecated..) and present it in a useful front end. For example, a restaurant review site that will let you search by locations and tastes and find all restaurant with >3 stars rating, or an events site that let’s you search for events today in San Francisco etc.
There may be a cleaner alternative to the text I’m proposing. In any case, the key here is to implement microformats searchability on the publisher’s side. The downside to this solution is a less than elegant implementation, the upside is minimal effort required from the publisher’s side, and none at all on the consumer side. I think this may be a good step towards creating microformats-based services, which in turn may encourage more content publishers to use microformats.