Search engines today usually have a proprietary algorithm for ranking their results based on how many links point to the result, how many search terms the result has, and other factors.
The problem with this approach is that once calculated, the rankings do not change. For example, in searching for jaguars we may find sports teams, cars, or the animal. How do we focus on the one particular idea that user has in mind? This is a central problem in searching.
iMetaSearch uses a new technique called dynamic relevance to help solve this problem. The key is that relevance scores are not just calculated once, they change depending on feedback from the user about what results are relevant. You can either click clusters, mark results, or mark index words that you are interested in, and all result and word relevance bars are re-calculated. From there, you can easily find relevant results.
Related Topics:
About Results