<div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Good point, and there is other information in the VBI (or its digital<br>equivalent) in addition to the rating info, such as closed-
<br>captioning. A lot of commercials, even those that are not captioned,<br>issue a "clear" command at the start to make sure there is no garbage<br>on-screen left over from the program or preceding commercial. That
<br>would help in identifying "blocks", even if it didn't help in<br>deciding if that "block" was a commercial. For example a block that<br>was exactly 30 seconds long would have a higher "score" if it started
<br>with a clearCC signal.</blockquote><div><br> </div><br></div>
Hmm. I suppose we could route these questionable caption blocks thru spamassassin and use the bayes filter to help determine, based on the use of words, the "score" of that block of video.<br>