for http://www.latimes.com/business/la-fi-ct-nintendo-20110608,0,6032667.story

the api returned "author"=>"Los Angeles Times"

Is that because it uses the meta information as the first source?

In the content, there is By Alex Pham and Ben Fritz, Los Angeles Times which in our parsing code we pick up. We assume that a "byline" will override the values in the header, since we have found the header to many times incomplete data.