Not sure I see the point of Google's blog search, either

by Michael S. Kaplan, published on 2005/09/14 12:59 -04:00, original URI: http://blogs.msdn.com/b/michkap/archive/2005/09/14/466217.aspx

Though it looks like /. is having a great time with the idea. Thought I have to admit the funniest comment I saw came from hyperstation, who said:

Which probably says more about the blogsophere than Google's attempt to index it! There are a lot of people writing blogs who should instead spend more time reading them. If you know what I mean. If you don't, then I do plan to blog about this one day, I just have not quite gotten to it yet....

Of course, the Google blogging index is not all there yet. So their effort is actually more like trying to index a litter box, turd by turd, and missing most of them.

I think that finding blogs that deal with subject material you need or find interesting is useful.

The question is how to put all the information presented by blogs in a coherent form so people can use it to find others who have similar problems, etc. Maybe someone has some great ideas that would benefit your work and you don't know they exist.

Now, traditional search doesn't really meet any of these goals. Despite the fact that all venders claim to have excellent search technology, I often have to look at 100s of pages using search before I get the information I want. With blogs, this will be much much worse to the point where search just doesn't work. There should be ways to make it work though..

For example, I could use heuristics to find the experts on any given subject in a given community. I could then look at which pages they commonly look at, comment etc, to see who they feel are experts in the same material. I figure out weighting criteria to determine additional experts and then refine my set. Note that the word Index is missing from this. Looking for words on a page is highly unlikely to get me to the right people instead, I am looking at web traffic patterns.

Eventually, I might be able to do type in localization and get "Sorting it all out" and then some of your friends web pages afterwards. If I was looking for a particular portion of localization like Windows Vista Localization, the search would again analyze traffic patterns, find the "Windows Vista Localization" experts and then find which subset of their pages deal with the specific topic weighting them by page hits or comments.

Now, this would be useful.

While I have to agree, diary blogs have no value outside of 1st person and your immediate friends (if they even care that much).

But there are many usefull blogs out there.

And on a slap to Microsoft, if they don't hurry up and improve their blogs.msdn.com search, Google will be my engine of choice for searching their blogs, soon.

I don't believe Google is creating a blog search because it is very useful in of itself, but rather to improve their more conventional web search. Creating a blog search may provide Google with some knowledge and expertise that will help them reduce blog's overwhelming influence on PageRank and thus reduce their susceptibility to Google bombing.

Having some sort of search service with blog in the name probably also serves some marketing purposes regardless of its technical merit.