|
Negative Search is the elimination of information which is not relevant from a mass of content in order to present to a user a range of relevant content. Negative Search is different from both Positive Search and Discovery Search. Positive Search uses the selection of relevant content as its primary mechanism. Discovery calculates relatedness (between user intent and content) to present users with relevant alternatives of which they may not have been aware. Negative Search applies to those forms of searches where the user has the intention of finding a specific, actionable piece information but lacks the knowledge of what that specific information is or might be. Negative Search can also apply to searches where the user has a clear understanding of Negative Intent (what they don't want) rather than what they do. Examples of Negative Intent are: - Job searching: someone knows they want a new job but they have no idea what it might be. They just know what they don't want. - Online dating: someone is looking for a dating partner, but cannot identify what criteria they are looking for. They just know what they don't want. - An investigator is looking for a car but has no other information on that car on which to base a search. ==Negative Search Classifiers== If there are two forms of search (positive and negative) it follows that there are two forms of classifier models: Inclusive Classifiers and Exclusive Classifiers. Countries of the World are a good example of a MECE list. A positive search for the country Kenya would identify content referencing Kenya and present it. A Negative Search for the country Kenya would exclude all content relating to other countries in the world leaving the user with content of some relevance to Kenya. 抄文引用元・出典: フリー百科事典『 ウィキペディア（Wikipedia）』 ■ウィキペディアで「Negative search」の詳細全文を読む スポンサード リンク
|