I left a long comment on an ERE thread realated to Monster's new search offering.

My organization has been intimately involved with resume search since the first hour of business. It's a complex subject involving multiple audiences and multiple uses for search results. There are multiple material unanswered questions involved:

- is there anything specific to resume search that speaks to different methodologies than those used in other mission-critical text search solutions ? For instance does Google use a specific method or not? And if not, why not? Is it something specific to recruiting?

- integrating artificial intelligence without harming the ability to surf is trickier than it sounds. Surfing is a metacognitive strategy that depends on the end-user changing a single variable (the search string) and comparing the differing results when run against the same data set. When the search engine applies its Mojo, the end-user has no way to determine how the changed search terms change the result set: it's a black box. Not being able to surf properly is a serious comprom