The anatomy of a large-scale hypertextual Web search engine
WWW7 Proceedings of the seventh international conference on World Wide Web 7
Improving web search results using affinity graph
Proceedings of the 28th annual international ACM SIGIR conference on Research and development in information retrieval
Tracking news stories across different sources
Proceedings of the 13th annual ACM international conference on Multimedia
Fast tracking of near-duplicate keyframes in broadcast domain with transitivity propagation
MULTIMEDIA '06 Proceedings of the 14th annual ACM international conference on Multimedia
Integration of association rules and ontologies for semantic query expansion
Data & Knowledge Engineering
Integration of association rules and ontologies for semantic query expansion
Data & Knowledge Engineering
Storyline-based summarization for news topic retrospection
Decision Support Systems
Multi-document summarization using cluster-based link analysis
Proceedings of the 31st annual international ACM SIGIR conference on Research and development in information retrieval
Biased LexRank: Passage retrieval using random walks with question-based priors
Information Processing and Management: an International Journal
Key image extraction from a news video archive for visualizing its semantic structure
PCM'04 Proceedings of the 5th Pacific Rim conference on Advances in Multimedia Information Processing - Volume Part I
Multimodal News Story Clustering With Pairwise Visual Near-Duplicate Constraint
IEEE Transactions on Multimedia
| Hi-index | 0.00 |
Pseudo-relevance feedback is a popular and widely accepted query reformulation strategy for document retrieval and re-ranking. However, problems arise in this task when assumed-to-be relevant documents are actually irrelevant which causes a drift in the focus of the reformulated query. This paper focuses on news story retrieval and re-ranking, and offers a new perspective through the exploration of the pair-wise constraints derived from video near-duplicates for constraint-driven re-ranking. We propose a novel application of PageRank, which is a pseudo-relevance feedback algorithm, and use the constraints built on top of text to improve the relevance quality. Real-time experiments were conducted using a large-scale broadcast video database that contains more than 34,000 news stories.