After a few hours of testing Google Blog Search (GBS), here are a few comments, complaints, and things I would like to see in the future.
- Google needs to do a better job explaining what Google Blog search indexes vs. what’s available from Google News.
- The Google Blog Search “related blogs” section section offers access to non-blog RSS feeds like weather info from NOAA. Is the NOAA home page a blog? Please explain. Plenty of valuable and current info, in addition to blog content, is available via RSS and Google Blog Search could do a lot to promote non-blog uses of the technology.
- Explanation that GBS is NOT always a full text search. In other words, if it’s just a “snippet” feed, you might be missing useful material. Actually, other blog/rss engines should do the same thing and explain (aka teach) users what they are and are not searching. If it’s older material, Google could also provide a link to check Google’s main web index where the complete blog MIGHT be part of a frequent crawl/recrawl.
- I’m not a fan of filters but when you promote the fact that you offer one (it’s mentioned in a guide to GBS that Google sent me) it should work. I searched (Wed. afternoon) for words that are blocked on Google Web Search but they’re currently not blocked when searching the new blog database. I activated SafeSearch (strict filtering) from the preferences page and from the advanced search page and still no go.
- Ability to combine other Google Blog Search syntax with the link: command. Btw, combining other syntax with link: is also not possible with Google Web Search.
- Option to remove from results blogs that simply scrape headlines from other blogs especially when the blog originally posting the item is included.
- Option to quickly add blogs discovered via Google Blog Search to your aggregator, personalized page, or Google Sidebar with just a click. Something similar for pre-emptive search queries delivered via RSS would also be a timesaver.
- A standalone blog directory organized by topic(s) based on publisher suggestion and mining the text of the blog over a period of time.
- Option to limit by location of publisher if a Geotag is used. BlogDigger Local is offering this type of thing.
- Clustering of “related” blog posts linked underneath each result. A great way to find related posts.
- Improved Relevancy: One example. I did a search for my name [Gary Price] and found several non-relevant results in the first 20. In the second result “Gary” is found in one part of the post and “Price” in another. Then, back to results mentioning yours truly before posts discussing the “price” of movies for sale. Note: Yes, of course, I could search “Gary Price” and get more precise results but most searchers don’t use quotes to denote phrases.