|
Ironically, the idea of using an adaptive spam filter for literary pattern recognition is not entirely without merit (although the results would be more meaningful if they started with a scene-based structural analysis and gave a percentage-weighted interpretation, based upon similarities to multiple authors).
As a quick test, I fed the site a selection of scenes from two of my unpublished stories. Each time I offered the same scene, the site said it resembled the same author... but out of a dozen different scenes I tested, I was never compared to the same author for differing scenes.
If I wasn't a writer, and was instead looking for a way to kill time, I might even try to develop the concept further. It has potential, if you aren't tied to a serious interpretation of literary worth.
- M.
|