Shaping-Up Multimedia Analytics: Needs and Expectations of Media Professionals

Abstract

Abstract. This paper is intended to help clarifying what multimedia analytics encompasses by studying users expectations. As a showcase, we focus on the very specific family of applications doing search and navigation of broadcast and social news content. This paper is first describing what professional practitioners working with news currently do. Thanks to extensive conversations with media professionals, mockup interfaces and a human-centered design methodology, we analyze the perceived usefulness of a number of functionalities leveraging existing or upcoming technologies. This analysis helps (i) determining research directions for the technology underpinning the very recent field of (multi)media analytics and (ii) understanding how multimedia analytics should be defined. In particular, dependency to the domain is discussed: are multimedia analytics tasks domain-specific or can we find general definitions?

Stay Tuned! An Investigation of Content Substitution, the Listener as Curator and Other Innovations in Broadcast Radio

“…This section will extensively yet concisely review the media experience innovations that jointly have shaped our work. Please note that our work is situated exclusively in an entertainment or leisure setting; radio consumption as part of professional scenarios (e.g., [25]) is out-of-scope. Also out-of-scope are technological innovations in the radio domain (e.g., DAB Multimedia Object Transfer [17] or RadioDNS [18]); as stated in Section 1, our work focuses exclusively on the conceptual rather than technological level.…”

Linking Multimedia Content for Efficient News Browsing

“…is allows to fuse clusters that deal with related events, e.g., the continuation of a story, thus reducing their number [1]. According to user studies, the temporal relation between news items is the most important to media professionals [7], allowing them to reveal the causes and consequences of some events. However, similarly to clustering, when considering large timeframes, it becomes hard to display the list of news stories that are available in the dataset.…”

“…e temporal relation is clearly the most important type of relations according to media professionals [7]. But it is insu cient alone, in particular when exploring large news datasets that include articles with very similar content from di erent newswires that tends to clu er timelines.…”

Language-based Construction of Explorable News Graphs for Journalists

“…In this view, the user can initially see the entry point document itself (A) and the links that departs from it. In addition to the original content, metadata and keywords are displayed (B), as both were judged crucial in the preliminary usefulness studies (Gravier et al, 2016). Links appear in one of two ways.…”

“…News data depend a lot on chronology, which resulted in many approaches organizing collections as timelines so as to be able to follow the evolution of specific stories. The temporal relation is clearly the most important type of relations according to media professionals (Gravier et al, 2016). But it is insufficient alone, in particular when exploring large news datasets that include articles with very similar content from different newswires that tends to clutter timelines.…”