Except you don't know you don't want to read it if you're using something as crude as looking for a string in a headline. I maintain that blindly ignoring content that has a single mention of a particular term is sticking your head in the sand (but everyone's entitled to interpret the situation as they like!).

Your fictional app would likely rely upon categorization as an indicator. On news sites, content managers determine the interest areas for a story and you can then filter them. Stories that are really important will spill over to other categories (such as, on most news sites, the "main"/general category) and stories that seem to be about one topic but are really about another should, ideally, be categorized appropriately (e.g. "iPad Designer Jonathan Ive Dies" isn't an iPad story - crude example, but just to make a point).

Filtering out items based on a string of letters in a headline (as suggested in this case) is not as elegant or robust. You might filter out "football" but still be interested if the plane carrying the Dallas Cowboys went down wiping out the team. You might filter out "health care" to avoid political discussions on HN but miss out on a story about a healthcare entrepreneur. It's as crude as early spam filtering. (And like spam filtering, could be improved by software that analyzed the content of the destination pages.. but relying on the headline alone is too flaky.)

As a temporary measure to filter out a glut of stories (as with the iPad affair), it can work, but as a mid/long term strategy, filtering by headline is crude and prone to covering up stories you might actually find interesting. The quickest and most precise filtering device out there is the one connected to your eyes.