Suggestion: weighting recommendations option

I sometimes get recommendations for books by authors that I haven't read in years. An example in this month's list is Rhys Bowen. I have a column (or data field) in my library for "date read" (=date finished). Can there be an option for the recommendation software to use this data to weight the recommendation by temporal distance from today? What I mean is, if there was a recommendation for a book by an author whom I haven't read in 20 years, it would put it much further down the result list than a book by an author that I last read five years ago, let alone one this year. A recommendation for an author that I last read in 1985 isn't very useful.

It could also use that data to weight by subject in a similar way, but I can't presume to know how easy this would be to make use of or program into the recommendation software.

If this is implemented, I'd like it able to be toggled off. I have authors I read many years ago that I don't remember that I'd like to read again.

I'm with >2 gilroy:.

I like this idea a lot. Most writers I liked, I liked well enough to stick with so filtering this way would make it easier on me.