David Brain writes,
"in this years’ Edelman Trust Barometer we asked opinion leaders in 18
countries about the news sources they rely on most for information on
companies. Not surprisingly, local sources dominated, but three global
brands appeared regularly among the local names. Most of us could
probably have guessed the BBC and CNN would be mentioned, but the
number of respondents ticking the box marked Google confirms that the
concept of ‘search engine’ and ‘news source’ are becoming less
distinct. Last I looked, Google did not employ a single reporter."
[The graph on the original post proves interesting and is certainly worth a look.]
Google, Brain points out, has recently launched a local news service. The post announcing the new service explains how Google does it:
"We’re not simply looking at the byline or the source, but instead we
analyze every word in every story to understand what location the news
is about and where the source is located."
A look at the Google News Local page for St. Albans, a small City in the county of Hertfordshire, England, pulls up some relevant content but is also quite noisy, with the algorithm also pulling in headlines from West Virginia and Vermont USA and Australia. Useful then, I suppose, if you happen to be searching for news on a location with a unique place name but otherwise not so great.
There’s also, hiding over on the left side of the page, a new location specific blog search. The top results for St. Albans were spot on but further down there’s the same problem – the algorythm seems unable to distinguish between multiple cities named St. Albans and, worse, picks up lots of classified advertisements, newspaper articles and other non-blog content.
It’s a good idea but, for me at least, not good enough to be very useful for locality specific searches – and serves as a reminder that fancy algorithms can’t yet match human editorial judgement.