With search becoming increasingly verticalized, I am seeing both the need and demonstration not just of searching within a specific domain, but also ranking results along multiple dimensions in that domain. For example, Indeed lets me search for positions with certain job titles, as well as identify how recently those jobs have been posted. Blinkx.tv has a great slider-bar feature that lets me shift between the importance of date and relevance in ranking my results. It even lets me toggle the sources of video content. Kayak has similar slider-bars allowing me to sort through flights easily by time-window.
But I (and mabye others?) want even more. As we become more sophisticated in what we are searching for, we should be able to toggle and sort though the dimensions of what is most important to us. For blog searches, sometimes I want to ensure that I read something with strong authority, sometimes I want to read something with the utmost timeliness, and sometimes I want to read something with highly dense mentions of a certain topic.
Google’s PageRank philosophy worked well in identifying the most relevant static pages. But as we are continuing to move towards the Incremental Web, perhaps the criteria for determining the best given search result is multidimensional rather than one-dimensional – and easily changeable by the user.