Term-weighting

• Definition – The assignment of numerical values to distinguish the “value” of document (and query) descriptors, more commonly known as “index terms.”

• Example – “History of England” versus “English Language”

  • Perhaps “England” is a descriptor of both works
  • Since the history of England presumably has more to do with England than the English language does – the latter would likely be more concerned with linguistic properties than the nation of England – one would want to assign a higher weight value to the term for the former document