The Scientist :: The Future of Citation Analysis, Oct. 24, 2005: "The Future of Citation Analysis
The challenge is to track a work's impact when published in nontraditional forms"

It's intriguing to think that computers are getting better at parsing text better, but unsurprising that people are still needed to actually work out what makes a citation or acknowledgement valid.
I wonder what the citation/acknowledgement relationship or ratios are in different fields of science ... would 'softer sciences' have a greater ration of citations to acknowledgements than 'harder sciences'? A thought to be examined at a later date, I think! (And preferably by someone with a much better grasp than I currently have on statistical computation!)