Google has just updated its long-neglected blog search engine with some new functionality -- categorization by topic area, and grouping blogs around "hot" news topics. A few people are already trotting out the possibility that this is a "Techmeme Killer," but the reality is Gabe Rivera's blog indexing and aggregation service is still far stronger in all areas save one.

On the surface of it, Google Blog Search is greatly improved over the version that I first started using in 2005. There are 11 categories, many in media and technology, and the default presentation groups individual posts by hot news topics. Today they are heavily weighted toward the vice presidential debate (Gwen Ifill controversy, Sarah Palin vs. Joe Biden, etc.).

Sounds great, right? But dig a little deeper, and Google's long-term problems with indexing blogs bubble to the surface. The secondary results and sort-by-date search results are filled with scraper blogs, wrapped in AdSense and affiliate ad blocks. In other words, spam.

You don't have to dig hard to find it. The splog operators tune their blogs to scrape content based on hot search keywords, and this content is indexed by Google and Google blog search. This morning, the second blog in the "Apple drops NDA for iPhone app developers" block was a Hong Kong scraper, and the first search result for "nintendo dsi" was a Russian scraper that threw off very unsafe-looking pop-ups. This isn't what people want to look at.

Techmeme occasionally indexes people reprinting press releases or merely repeating news without any commentary, but the top-level items are usually interesting and relevant. And while Techmeme's A-list emphasis is sometimes frustrating, it is a better indicator of quality than Google's algorithms.

The one area in which Google blog search trumps Techmeme is audience. Google can drive far more traffic to the front page of Google Blog Search and integrate results into Google News and organic search. That would normally be the makings of a hit service and maybe even a Techmeme killer. But until Google's PhDs figure out a way to effectively separate the wheat from the chaff, I don't think many visitors will be impressed with this product.

» posted by ITworld staff

The Industry Standard

• Email
• Print
• Share
  • Slashdot
  • Digg
  • Stumble
  • Reddit
  • Newsvine