| 
|
Video demonstration from the 1998 CHI conference. One method for locating information on the World Wide Web is to use a search engine (SE). Given a set of terms, a SE will return a list of documents containing those terms. Often though, this list of documents is extremely large. Unfortunately, there are currently no tools to assist the information seeker in determining whether these documents contain desired information, or just submitted terms. Two types of SE errors are possible: false positive errors result from the many connotations which words may convey, and false negative errors result from different wordings that express similar meanings. To solve these difficulties, we focus on meaning rather than &QQ, developing a technique called Semantic Highlighting. Author(s): Ali Hussam, Brian Ford, Jack Hyde, Ali Merayyan, Bill Plummer & Terry Anderson |
MPEG-1 • 121.80 MB 
