This piece of news is all over the shop this morning:
Of course, patents don’t mean such a system exists, or will do so in future. However, if Google do use this patent, it may work like this:
- determine whether the search query corresponds to at least one query theme of a group of query themes
- is the page listed in a known directory, and/or have a relevant host name
- combine these factors to produce relevant results
The patent states the aim of the invention is to “enhance the ranking of search results by integrating editorial opinion”. The patent describes using themes (grouping of related terms) and hierarchical directories.
“For example, if the query theme is “sites that help in finding accommodation,” then web hosts listed under the Open Directory category “http://dmoz.org/Recreation/Travel/Lodging” can be taken as favored sources”.
Much of this patent appears to deal with the further validation of on-page content by checking against other criteria, namely directory listings and word themes.
“For the query theme “sites that provide free software downloads”, an exemplary rule may be that the search query must contain the words “free” and “download” and an exemplary directory topic may be “Computer: Software: Shareware.” “
Much of this stuff has been speculated about for years, however it’s interesting to see it written down in “official” form.