NAVIGATING VIDEO SCENES USING COGNITIVE INSIGHTS

Embodiments of the invention include method, systems and computer program products for obtaining a scene from a database. Aspects of the invention include receiving a search request for a scene stored in a database including annotated video content. One or more keywords are extracted from the search request. Extended keywords are generated for each keyword. The extended keywords and keywords are compared to the annotated video content to identify target scenes including target scene annotations. A confidence rating is assigned to each of the target scenes based at least in part on a correlation between the target scene annotations for each of the target scenes and the extended keywords and keywords. And at least one of the target scenes are displayed to the user, wherein the at least one of the target scenes is determined based on the confidence ratings.