Research has shown that the average user's search query is only 1.2 words long. Short Boolean queries cannot express exactly what the user is looking for, and as a result, search results returned by the engine are often irrelevant and incomplete. Automatic categorization can solve this problem, without requiring that users become trained researchers.

Problem				Solution
Users search for subjects. Users searching for documents about computers are likely to use the search term "computers." But the name of the subject may not appear in the document at all. For example, a story about IBM's new product may not contain the word "computers," because it is obvious that computers are involved.				Concept searching: With automatic categorization, documents have meta data that defines what they are about. The search engine can then find stories about "computers" even if they don't actually contain that word.
Stories containing synonyms are not returned. A 1.2 word query can't contain all the possible synonyms of the terms being searched. If the user searched for "computers," stories containing the word "PC" won't be returned at all, and the user will not know they exist.				Stories containing synonyms are in the same category. With automatic categorization, all the documents about computers will have the "computers" category as meta data. Stories containing the word "PC," or another related term, will be retrieved because they are in a relevant category.
Irrelevant search results. In the absence of meta data, search engines must rely on the occurrence of words in a story to rank its relevance. A search engine may give a low ranking to a story that is actually about the subject in question, simply because it contains too few occurrences of the search term, and bury it on the 17th results page.				Ranking by meta data: With automatic categorization, search engines can rank documents higher if they are actually about the subject, or contain related keywords: These are the stories that will have real value for readers. Other documents, which contain the search term but do not belong to the relevant category, will be ranked lower.
Users never get the "big picture." Search results are always missing something, because most search queries aren't perfect. Users never know for sure that they are seeing all the items about the subject at hand.				Comprehensive search results: With automatic categorization, documents are consistently classified into categories. A user searching for a certain category's concept will get all the stories belonging to that category.