Using GA Event Tracking to create learning search

I am wondering if there is a way to use Google Event Tracking to watch user behavior after a search results page is delivered and then feed that data back into my search algorithm to improve the ordering of that page the next time it is rendered?

For example

User searches 'Blue Widget' - I have 50 of them and I display them in some order, say best selling first.

User looks at the 50 results and clicks on the Widget A which is in position 10.

I want to capture that interaction and use it to see whether Widget A should be moved up to position 5 (for example) the next time someone searches Blue Widget

Further, the user who clicked on Widget A then bounced back to the results page - black mark for Widget A, relegate it back to position 10 (for example).

Or. User doesn't bounce back, but instead adds the product to their bag. Big tick for Widget A, move him up to position 3 (for example).

User goes on to complete the order, another tick.

User exits site from Widget A page, black mark.
User refines the search, black mark for all Widgets shown on the results page

Etc.

Is there some way to use events to grab this data and store it alongside the search terms AND the Widgets that were shown?

Thanks...