Video-on-demand （VoD） services have become popular on the Internet in recent years. In VoD, it is challeng- ing to support the VCR functionality, especially the jumps, while maintaining a smooth streaming quality. P...Video-on-demand （VoD） services have become popular on the Internet in recent years. In VoD, it is challeng- ing to support the VCR functionality, especially the jumps, while maintaining a smooth streaming quality. Previous stud- ies propose to solve this problem by predicting the jump tar- get locations and prefetching the contents. However, through our analysis on traces from a real-world VoD service, we find that it would be fundamentally difficult to improve a viewer＇s VCR experience by simply predicting his future jumps, while ignoring the intentions behind these jumps. Instead of the prediction-based approach, in this paper, we seek to support the VCR functionality by bookmark- ing the videos. There are two key techniques in our pro- posed methodology. First, we infer and differentiate view- ers＇ intentions in VCR jumps by decomposing the inter- seek times, using an expectation-maximization （EM） algo- rithm, and combine the decomposed inter-seek times with the VCR jumps to compute a numerical interest score for each video segment. Second, based on the interest scores, we pro- pose an automated video bookrnarking algorithm. The algo- rithm employs the time-series change detection techniques of CUSUM and MB-GT, and bookmarks videos by detecting the abrupt changes on their interest score sequences. We evaluate our proposed techniques using real-world VoD traces from dozens of videos. Experimental results suggest that with our methods, viewers＇ interests within a video can be precisely extracted, and we can position bookmarks on the video＇shighlight events accurately. Our proposed video bookmark- ing methodology does not require any knowledge on video type, contents, and semantics, and can be applied on various types of videos.展开更多
文摘Video-on-demand （VoD） services have become popular on the Internet in recent years. In VoD, it is challeng- ing to support the VCR functionality, especially the jumps, while maintaining a smooth streaming quality. Previous stud- ies propose to solve this problem by predicting the jump tar- get locations and prefetching the contents. However, through our analysis on traces from a real-world VoD service, we find that it would be fundamentally difficult to improve a viewer＇s VCR experience by simply predicting his future jumps, while ignoring the intentions behind these jumps. Instead of the prediction-based approach, in this paper, we seek to support the VCR functionality by bookmark- ing the videos. There are two key techniques in our pro- posed methodology. First, we infer and differentiate view- ers＇ intentions in VCR jumps by decomposing the inter- seek times, using an expectation-maximization （EM） algo- rithm, and combine the decomposed inter-seek times with the VCR jumps to compute a numerical interest score for each video segment. Second, based on the interest scores, we pro- pose an automated video bookrnarking algorithm. The algo- rithm employs the time-series change detection techniques of CUSUM and MB-GT, and bookmarks videos by detecting the abrupt changes on their interest score sequences. We evaluate our proposed techniques using real-world VoD traces from dozens of videos. Experimental results suggest that with our methods, viewers＇ interests within a video can be precisely extracted, and we can position bookmarks on the video＇shighlight events accurately. Our proposed video bookmark- ing methodology does not require any knowledge on video type, contents, and semantics, and can be applied on various types of videos.
