A learning agent that assists the browsing of software libraries

摘要: Locating software items is difficult, even for knowledgeable designers, when searching in large, complex and continuously growing libraries. This paper describes a technique we term "active browsing". An active browser suggests to the designer it estimates be close target of search. The novel aspect browsing that entirely unobtrusive: infers its similarity measure from designer's normal actions, without any special input. Experiments are presented which system succeeds 40% time identifying before has found it. additional experiment indicates this approach does, indeed, speed up searches.