For many designers & graphic artists, a first experience with disambiguation will be during an image search at a site like Getty One or Corbis. We’ll conduct a search for “marine” and a page will return asking us to clarify if we meant the soldier or the ocean. This is an example of a disambiguation page, a clarifying feature designed to interpret the meaning of the query.
Homographs (those words which are spelled the same, but have different meanings) often require disambiguation in search because there is no additional contextual data to extract a meaning. This concept, however, extends further into the world of search as the engines attempt to use methods to automatically disambiguate queries in order to provide relevant results. There are currently several methods under study for conducting disambiguation:
- History – Using a searcher’s visit history and search history, a search engine could conclude whether the searcher is seeking information on one topic or another.
- Geography – Using IP-location (geotargeting), a search engine can predict what is the more likely relevant subject for a given query.
- Syntax & Context – The way in which a word is used or the words accompanying it may give the search engine clues as to the searcher’s intent.
Disambiguation is an important topic for SEOs, especially those involved with search terms or phrases that may have multiple meanings or whose searchers may be seeking different types of information. As I build pages targeted to these types of terms, I always seek to add classification information that defines the pages subject through context and meaning. For example, if I were writing a page on the subject of Ray Charles’ song, Georgia, I would also want to provide a page or links to pages on other topics to which the term Georgia is relevant, as Wikipedia’s Georgia disambiguation page does.
