Multimedia content is presented to a user. Questions presented by the user are monitored. A core concept of the question can be identified. The multimedia content can be analyzed to determine whether the multimedia content contains a possible answer to the question. If the multimedia content contains a possible answer to the question, but the possible answer is covered in a plurality of segments of the multimedia content, a confidence score can be assigned to each of the plurality of segments that cover the possible answer. The confidence score can indicate a probability that the respective segment answers the question. Based on the confidence score assigned to each of the segments, a segment for which there is a high level of confidence that the segment answers the question can be selected. Based on the selected segment, an answer to the question can be automatically generated and presented to the user.