The article discusses how search engine algorithms such as PageRank may not be effective for searching blog contents. PageRank works by assigning a value between 0 and 10 to represent the “weighting” of each webpage. The value, calculated from the number of in-flowing links and the popularity or authority of those links has been successful in yielding satisfying search results for ordinary webpages. However, the method fails miserably for blogs because it is extremely hard to qualitatively judge the links connecting to blogs. Another problem is that blogs generally receive fewer incoming links regardless of its actual number of readers. To solve this problem, Apostolos Kritikopoulos and his friends at the Athens University of Economics and Business in Greece have devised a system similar to Google’s PageRank, called BlogRank. BlogRank attempts to rank weblogs by exploiting tags and the contribution made by authors and commenters and the way in which these individuals add content to more than one blog.  

Weblogs are becoming increasingly popular and much of the information online is being stored in the blog servers. It is important to come up with a way to optimize searching so that these information stored in blog posts can be more easily accessed. As we’ve seen in the class lectures and homework assignments, incorrectly “ranking” blog pages can potentially cause huge changes in the “weight” scores of each webpage. A blog may have millions of subscribers or readers while having very few incoming links. However, BlogRank may not be as easy to implement due to the nature of weblog pages. For example, it would be difficult to constantly update and get hold of the full list of commenters on many blogs, and achieving their identities may also prove to be a hard task. Nevertheless, successful implementation of the algorithm will most likely yield much improved search results for information gathering on the web. http://www.technologyreview.com/blog/arxiv/23252/ http://arxiv.org/abs/0903.4035v1