Posts Tagged ‘blog search engine’

The case for blog search engine

August 30, 2008 4 comments

I see more and more people raising the questing about the need for blog search engine, especially when Google is doing such a great job finding good content from blog as well as from web sites. It seems like that Google itslef is not investing too much in their blog search too. So, in this post I will explain what I think should be the duties of a blog search engine and why I still see a need for one.

Blog search engines (should) serve multiple purposes

  • Finding great bloggers, blogs and blog posts
  • Recognizing great bloggers, blogs and blog posts – rank.
  • Categorizing blogs and bloggers in multiple ways not limited to content type. Categorize blogs by their objectives: personal blogging is not the same as corporate blog or professional bloggers, subject expert, politics, go green, artist or others. It is not just about what that the blogger writes about but also about what the blogger is trying to achieve.
  • Monitoring blog and blogger progress – is this blog alive? a shooting star?
  • Web-now – see Twitter Search Trending Topics, Twingly’s Hot right now or Technorati’s what’s percolating in blogs now
  • Alerts – a list of new blogs in a given category that are doing well
  • Community building – increasing cooperation among bloggers (e.g. you should read this blog)

What do we need to know?

  • The top bloggers in a category
  • The top blogs in a category
  • The top blog post in a category

Who needs it?

  • The readers – to know what to read, what is going on in real-time
  • The blogger
    • To present a case to a sponsor
    • To know whom to look up to
    • To see and share about the blogger progress
  • The business
    • To know where to buy ad real-estate or whom to sponsor
    • PR – where to spend my effort effectively

The challenges of blog search engines today.

Using the reaction counting method for ranking, the service needs to distil humans actions from automated (bots) one in order to be accurate. So far this is not working well and adds another questions mark around the validity of blog search engines.

Here are some example for both:

  • Human reactions
    • Blog post reacting to another
    • Update on Twitter or Jaiku
    • Digging on Digg
    • Submitting to social bookmarking site
    • Posting on a social network
    • Bloggers community
  • Bot reactions

The number of sites that offer posting of human blog reactions is growing faster than the crawling capabilities and sometime does not offer access to crawlers.

The service should also remove the “me” links from the count i.e. links from all the social object under the same owner.

A couple of thoughts

Maybe someone could think about another way to rank blogs and bloggers. Measuring traffic is probably a more accurate way (Alexa). The traffic is relative to the category. I assume that a blog about Technology will get more traffic than a blog about biology. The rank should be within a category and not across all (or not just across all blogs).

In my opinion there is a need for blog/blogger search engine but the emphasize of the search capability should be less around finding content (leave that to Google) and more about discovering leading blogs and bloggers. 

It does not need to be a free service at least not for the business. The premium or a sponsor account model could work as well.

Taking the Search personally

Are we going to see a new alternative added to the growing list of online content search solutions (I wrote about some of them here)?

So what do we have so far? popularity search engine (Google), semantic search engine, location aware search engine, blog search engine, spam free blog search engine, social search engine, conversational search engine, visual search engine, passive search engine, social bookmarking search engine. Did I miss any?

The new alternative is a profile based search engine. I’m not talking about Google Web History solution for personalized search engine using historical searches for better qualifying search result. It is a great idea though. I’m talking about using information, about me, from my profile, to better fit results to my search query. For instance if I’m a programmer searching using the Hibernate keyword I don’t want to see results about bears. If I’m a doctor querying about Viagra I don’t want to get all the spam in the world from now till eternity. If I’m 20 years old my world is different than the life of anyone else (if you call it a life after 30:)). Now these examples are for static profile attribute, yet there are many dynamic qualifiers that can be used to improve search results. One example was implemented already in location-aware search engines using your where about. Another is using your network dynamics like what that Delver is trying to do.  

It was the Web 2.0 applications with Facebook, as one of its leading example, that helped us to realize that one’s profile is not just name, email and contact. Your network (friends and groups), media choices (pictures, movies, books, and games), and activities (feed, expending your network, and participation) is also part of who you are. So, why don’t use it to help us find relevant content?


Get every new post delivered to your Inbox.