Assessing user behaviour in news video retrieval

Hollink, Laura; Nguyen, G.P.; Koelma, D.C.; Schreiber, A.T.; Worring, Marcel

doi:10.1049/ip-vis:20045187

Cited by 12 publications

(12 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The data in the transaction logs of experts and novices shows the experts made use of query-by-text less and visual strategies (query-by-image, query-by-concept) more, but the novices relied heavily on text search, despite poor quality and poor coverage of news with automatically detected on-screen text within the Visual-Only treatment [1]. Regardless of treatment, Full or Visual-Only, text search strategies dominated, as found by other TRECVID search investigators (e.g., [6,17]), but the excellent performance of experts when using all three query strategies led us to the goal of enabling increased use of these effective query mechanisms by novices as well.…”

Section: Baseline System (2004): Novices Ignoring Image/concept Querymentioning

confidence: 89%

“…Through the past five years, interactive retrieval systems evaluated in TRECVID have almost universally supported queryby-text, with that functionality responsible for most of the information retrieval success through TRECVID 2004 [5]. Query-by-image example is the next most frequently supported strategy across TRECVID participants [4,5,9], with query-byconcept not having success in early 2003-2004 trials [6,17] and not being implemented and tested as widely as the other query strategies.…”

Section: Video Search: Query By Text By Image and By Conceptmentioning

confidence: 99%

“…Thumbnails showing a key frame per shot arranged in temporal order in storyboards are typical of most TRECVID search systems (e.g., see [1,4,6,9]), with some extensions. For example, FX Palo Alto uses variable size thumbnails in a collage presentation [4] rather than fixed size, and MediaMill provides a "cross browser" for fast browsing of thumbnails in ranked results or timeline order in addition to a grid-based storyboard [14].…”

Section: Baseline System (2004): Novices Ignoring Image/concept Querymentioning

confidence: 99%

“…TRECVID interactive evaluations historically have emphasized only information retrieval effectiveness measures (e.g., mean average precision -MAP), and not other measures of end user utility. The vast majority of TRECVID interactive runs have been conducted by the researchers themselves posing as users [9], with a few notable exceptions [1,2,3,6,17]. Even these exceptions, though, use students as subjects rather than real world users.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Establishing the utility of non-text search for news video retrieval with real world users

Christel

2007

Proceedings of the 15th ACM International Conference on Multimedia

View full text Add to dashboard Cite

TRECVID participants have enjoyed consistent success using storyboard interfaces for shot-based retrieval, as measured by TRECVID interactive search mean average precision (MAP). However, much is lost by only looking at MAP, and especially by neglecting to bring in representatives of the target user communities to conduct such tasks. This paper reports on the use of within-subjects experiments to reduce subject variability and emphasize the examination of specific video search interface features for their effectiveness in interactive retrieval and user satisfaction. A series of experiments is surveyed to illustrate the gradual realization of getting non-experts to utilize non-textual query features through interface adjustments. Notably, the paper explores the use of the search system by government intelligence analysts, concluding that a variety of search methods are useful for news video retrieval and lead to improved satisfaction. This community, dominated by text search system expertise but still new to video and image search, performed better with and favored a system with image and concept query capabilities over an exclusive text-search system. The user study also found that sports topics mean nothing for this user community and tens of relevant shots collected into the answer set are considered enough to satisfy the information need. Lessons learned from these user interactions are reported, with recommendations on both interface improvements for video retrieval systems and enhancing the ecological validity of video retrieval interface evaluations.

show abstract

Section: Baseline System (2004): Novices Ignoring Image/concept Querymentioning

confidence: 89%

Section: Video Search: Query By Text By Image and By Conceptmentioning

confidence: 99%

Section: Baseline System (2004): Novices Ignoring Image/concept Querymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Establishing the utility of non-text search for news video retrieval with real world users

Christel

2007

Proceedings of the 15th ACM International Conference on Multimedia

View full text Add to dashboard Cite

show abstract

“…The clustered images are finally attached to the dictionary entry, and then attached to the OWL hierarchy. Polysemy in the hierarchy is dealt using query expansion [5]. For ambiguous terms, the associated entry in the leaf dictionary is formed of the term and its immediate hypernym.…”

Section: Image Clusteringmentioning

confidence: 99%

Using Semantic Commonsense Resources in Image Retrieval

Popescu

Grefenstette

Moëllic

2006

2006 First International Workshop on Semantic Media Adaptation and Personalization (SMAP'06)

View full text Add to dashboard Cite

Many people use the Internet is to find pictures of things. When extraneous images appear in response to simple queries on a search engine, the user has a hard time understanding why his seemingly clear request was not properly satisfied. If the computer could only understand what he wanted better, then maybe the results would be more precise. We believe that the introduction of an ontology, though hidden from the user, into current image retrieval engines would provide more accurate image responses to his query. Coordinating the use of an ontology (OWL representation of WordNet) with image processing techniques, we have developed a system that, given an initial query, automatically returns images associated with the query by specializing the query concept using only its deepest hyponyms from the ontology. We show that picking randomly from this new set of images provides a better representation for the initial, more general query. In addition, we exploit the visual aspects of the images for these deepest hyponyms (the leaves of WordNet) to cluster the images into coherent sets. In this way we can present the results in a structured, and even ontologically labeled, manner to the user.

show abstract