When undertaking complex search tasks that cannot be completed in a single session, resuming the search task can be challenging. When the search is conducted over multiple platforms, this challenge increases. While aggregated search interfaces support the synthesis of search results from multiple sources, cross-session support in such settings remains underexplored. In this research, we study a set of novel workspace approaches to this problem, within a digital humanities search context. We introduce an interactive cluster refinement (ICR) workspace that automatically organizes saved resources using thematic clustering, preserves their provenance, and provides an interactive method for refining the cluster to match the searcher’s mental model. Building on this foundation, we investigate two augmentations to facilitate reacquaintance with previous search activities: interactive passage highlighting (ICR+IPH), which lets searchers mark salient passages as a reminder of why resources were saved; and AI-generated summarization (ICR+AIS), which produces concise overviews of saved resources with inline citations. We combine these together into a fourth interface (ICR+IPH+AIS), where the AI-generated summarization is instructed to make use of the passage highlighting. We conducted a cross-session laboratory study with n = 36 participants to compare these interface alternatives with respect to measures of utility, perceived value, metacognition, exploratory search behaviour, and cross-session support. Both ICR+IPH and ICR+AIS improved outcomes relative to ICR. The combined ICR+IPH+AIS interface was most effective in enabling cross-session searching, yielding faster task resumption, higher resource precision, greater perceived value and knowledge gain, and stronger support for metacognitive planning and evaluation as well as exploratory search outcomes.

Interactive Information Retrieval (IIR) interfaces are typically evaluated using questionnaires that gather post-task subjective measures such as ease of use, usefulness, satisfaction, and user engagement, along with in-task objective measures derived from log analysis.However, a comprehensive evaluation requires a deeper understanding of user behaviour beyond such traditional measures.Integrating eye tracking data with logged feature use and emotional valence provides a multimodal approach to evaluating a search interface at the feature level.To validate this approach, we examined three search interfaces in a controlled laboratory study focused on exploratory search within the context of digital humanities archives.A key benefit of this multimodal approach is that it allows us to evaluate both traditional interaction with the search interface (looking at a feature, using it, and experiencing an emotional response) as well as passive interaction with the search interface (looking at a feature, choosing not to use it but possibly getting information from it, and experiencing an emotional response).Using this approach, we were able to identify specific features of the interfaces that generated positive and negative emotional valence responses when used, as well as features that generated such emotional valence responses when viewed but not used.Such feature-level assessments would be difficult to capture using other means, providing insight into the nature of the searchers' experiences using the search interfaces.

Abstract Searching across diverse information platforms, such as digital humanities archives, academic digital libraries, and encyclopedias, poses challenges in managing the queries issued to each platform and synthesizing the resources discovered. While search result aggregation interfaces address this problem, how best to present the search results from different platforms in the search engine results page remains an open question. In this research, we implemented three common approaches and developed a new technique for aggregating search results across three platforms: Europeana, our University's academic library, and Wikipedia. The three common approaches (1) use tabs to switch between the platforms, (2) interleave results from each platform producing a single list, and (3) use a bento box approach to group results from each platform. The new technique organizes the search results into thematic clusters irrespective of their source platform. We designed a controlled laboratory study using a within‐subjects design and exploratory search tasks conducted in the context of digital humanities searching. We collected data from 32 student participants, focusing on utility, perceived value, and diversity of saved resources. This study provides evidence that thematic clustering can be a beneficial aggregation approach, opening opportunities for studying different ways of representing and visualizing aggregated search results.

Search interfaces serve as the primary gateway to information retrieval (IR) platforms, with each IR platform possessing its unique interface. Yet, the challenge remains in delivering diverse content effectively to users, particularly when dealing with unfamiliar sources or content. Instead of platforms presenting search results from different sources in individual tabs for distinct sources, which can lead to user confusion and over-reliance on certain tabs, this paper investigates an approach to aggregate search results into a single, unified display, providing users with a consolidated list. Our research emphasizes novel presentation methodologies that combine insights from previous studies with advanced visualization techniques. The aim is to offer an intuitive and streamlined search experience for all users. Key research questions address: the design of interfaces to blend aggregated results while visually indicating their provenance; the advantages of such interfaces; the impact of search result diversity on perceived trustworthiness; and the applicability of the approach in structured data domains, specifically digital humanities archives and digital academic libraries. Our structured, iterative research methodology encompasses a three-phased approach: starting with low-fidelity prototyping, moving to medium-fidelity design iteration, and currently working towards functional prototype development. An upcoming controlled laboratory study, complemented by data collection tools like LogUI and eye-tracking, aims to gain a comprehensive understanding of user interaction and attention patterns and evaluate the proposed designs, providing insights into their effectiveness and the implications of visual result provenance.