Effectively detecting topic boundaries in a news video by using wikipedia
|Effectively detecting topic boundaries in a news video by using wikipedia|
|Author(s)||Kim J.W., Cho S.-H.|
|Published in||International Journal of Software Engineering and its Applications|
|Keyword(s)||Semantic interpretation, Topic boundary detection, Wikipedia|
|Article||BASE, CiteSeerX, Google Scholar|
|Web||Ask, Bing, Google (PDF), Yahoo!|
|Download and mirrors|
|Local copy||Not available|
|Remote mirror(s)||Not available|
|Export and share|
|BibTeX, CSV, RDF, JSON|
|Browse properties · List of journal articles|
Effectively detecting topic boundaries in a news video by using wikipedia is a 2014 journal article written in English by Kim J.W., Cho S.-H. and published in International Journal of Software Engineering and its Applications.
With the development of internet technology, traditional TV news providers start sharing theirs news videos on the Web. As the number of TV news videos on the Web is constantly increasing, there is an impending need for effective mechanisms that are able to reduce the navigational overhead significantly with a given collection of TV news videos. Naturally, a TV news video contains a series of stories that are not related to each other, and thus building indexing structures based on the entire contents of it might be ineffective. An alternative and more promising strategy is to first find topic boundaries in a given news video based on topical coherence, and then build index structures for each coherent unit. Thus, the main goal of this paper is to develop an effective technique to detect topic boundaries of a given news video. The topic boundaries identified by our algorithm are then used to build indexing structures in order to support effective navigation guides and searches. The proposed method in this paper leverages Wikipedia to map the original contents of a news video from the keyword-space into the concept-space, and finds topic boundaries by using the contents represented in the concept-space. The experimental results show that the proposed technique provides significant precision gains in finding topic boundaries of a news video.
- This section requires expansion. Please, help!
Probably, this publication is cited by others, but there are no articles available for them in WikiPapers.