Discovery work-space design to explore semantically aggregated graph-cluster spaces of multimedia document results
Umer Rashid, Abdur Rehman Khan · Behaviour and Information Technology · 2025
The advent of Web 2.0 allows users to generate multimedia content, leading to exponentially growing user-generated multimedia content. Subsequently, the users' information needs are transmuted from converse lookup to diverse exploration and discovery-oriented search activities. Despite being a gateway to accomplishing search tasks, search engines are inadequate in supporting the discovery of multimedia results. The conventional information organisation and presentation approach disperses multimedia content across various ranked positions and web pages, which causes a lack of relationships between the disjoint verticals, nonlinear navigation, and results space comprehension. Information discovery demands non-conventional search task support in seeking and exploration activities. The previous research disjointly discusses the factors supporting information exploration and discovery, which remain limited to theoretical discussion and have limited applicability to existing search engines. Therefore, in this research, we identify the key discovery-fostering parameters, aggregate them in a unified approach, and propose an end-to-end discovery solution with applicability in the existing search engine shortcomings to unveil the proposed approach's real-time comparability. Our proposed approach enables discovery via intelligible semantically aggregated multimedia graphs, visualisations, and Search User Interface (SUI) with a 95.26% success rate of cognitive walk-through. A between-study comparison (N = 44) of our approach and standard baseline revealed a 55.84% increase in usability (mean difference = 2.59, 95% CI [1.55, 3.63]) and a 73.61% increase in SUI satisfaction (mean difference = 6.01, 95% CI [3.76, 8.33]), both statistically significant at p<0.05.