作者: Tom De Nies , Erik Mannens , Rik Van de Walle , Wesley De Neve
DOI:
关键词: Information retrieval 、 Metric (mathematics) 、 Computer science 、 Task (project management) 、 Similarity measure 、 Semantic similarity 、 Context (language use) 、 Hyperlink 、 Similarity (psychology) 、 Jaccard index
摘要: In this paper, we describe our approach to the Search and Hyperlinking task at MediaEval 2013 benchmark. This focuses on video retrieval linking in context of a large rich dataset provided by BBC. Our makes use one three types audio transcripts, enriched with Named Entities. To compute similarity, adapt Jaccard metric results an unsupervised computationally inexpensive way searching multimedia content.