Matching

The goal of a lot of Search providers is to make the ultimate match between the person that queries and the most relevant result. The most relevant match has the highest value (e.g. the right advertisement at the right moment for the right person) and represents high monetization of content. There are at least two areas that need further exploration:

1. Pattern recognition: matching the features of content and the features of users, or intermediate metadata information (e.g. Opel driver reads Telegraaf with probability of .8) and
2. the algorithm that offers the most optimal user-content advice. Currently a lot of matching is taking place based upon word to word reference (e.g. article on banks, triggers advertisement on loans.) In most of the cases there is hardly any personal profile of the user taken into account, with the exception of dating sites.