Much of today’s focus on web search is on the challenge of relevant information retrieval from the World Wide Web. The rapidly growing number of new websites makes it difficult for search engines to return quality search results from the large number of queries it receives daily. What happens then, if we limit the problem of information retrieval to a smaller, yet closely knit subset of the web? In this post, I intend to explore the search engines of blogs.

The rise of Internet blogs has expanded into a tightly linked section of the web – a so-called ‘blogspace’ that has both content and structure. There are already many search engines dedicated specifically towards blogs, e.g. Blogdigger and Blogpulse. Yet how can we differentiate between a blog search and a typical web engine search? To answer this question, we must analyze blog search engine queries, and user behavior of blog searches. Studies of blog engines through search logs show that that majority of queries are informational in nature – i.e. the searcher is looking for knowledge about a topic. This differs from a typical engine search – where queries can also be navigational – i.e. the user is trying to find a specific site. The two main types of queries are contextual or conceptual. That is, searchers are either interested in blogs in which certain references are mentioned, or posts about a certain subject matter. In this sense, blog engines need to tailor to a much narrower and specific search range.

There are many similarities between blog searches and regular search engines. Users tend to be interested in only the top few results of blog engines. Hence, returning blog posts that are relevant is particularly crucial, since blog searchers are usually more attuned to their various interest areas in entertainment or politics. Hence the ideas of hubs and authorities are still applicable to blog search engines. Important bloggers whose opinions are deemed more reliable may have many links to them, while most bloggers will be part of a smaller network within the blog space. Although research into the characteristics of blogspaces is still being developed, there is no doubt, a necessity to bear user behavior in mind when designing future blog search engines.

References:

http://www.springerlink.com/content/j1254g6h75747725/