KnowledgeTrail: Generative Timeline for Exploration and Sensemaking of Historical Events and Knowledge Formation

TL;DR

KnowledgeTrail introduces generative timelines—dynamic, AI-driven interfaces that expand, contract, and reconfigure timelines in response to user curiosity. By co-constructing timelines of historical events and knowledge formation, KnowledgeTrail fosters breadth- and depth-based exploration, traceable relationships, and verification through citations. Two studies reveal that while generative timelines spark curiosity and serendipitous discovery, trust hinges on credible, configurable source citations and robust relationship explanations. The work argues for a new class of exploratory interfaces that balance serendipity with credibility, with implications for education, interface design, and AI-assisted sensemaking.

Abstract

The landscape of interactive systems is shifting toward dynamic, generative experiences that empower users to explore and construct knowledge in real time. Yet, timelines -- a fundamental tool for representing historical and conceptual development -- remain largely static, limiting user agency and curiosity. We introduce the concept of a generative timeline: an AI-powered timeline that adapts to users' evolving questions by expanding or contracting in response to input. We instantiate this concept through KnowledgeTrail, a system that enables users to co-construct timelines of historical events and knowledge formation processes. Two user studies showed that KnowledgeTrail fosters curiosity-driven exploration, serendipitous discovery, and the ability to trace complex relationships between ideas and events, while citation features supported verification yet revealed fragile trust shaped by perceptions of source credibility. We contribute a vision for generative timelines as a new class of exploratory interface, along with design insights for balancing serendipity and credibility.

Figures (10)

• Figure 1: How (A) Perplexity and (B) OpenAI's ChatGPT present their source citations — links to websites where the information is being sourced. Peplexity displays the (a) sources in a separate panel and also (b) in line. Similarly, ChatGPT uses the (b) in-line citation. We took inspiration from this design and applied it to KnowledegTrail.
• Figure 2: An example of event generation in KnowledgeTrail: Users can generate events related to a topic by (A) hovering over an event and (B) selecting the Events button in the Expand Bar or with the Search Bar (Fig. \ref{['fig:ContextualEventGeneration']}). The (C) arrows trace the path of generation to help track the exploration path. A description on how the events are related is then (D) displayed in the Side Panel. (E) Users are presented with (E & F) links to sources for verification. They can select (E) citation box or (F) inline link to open a new tab in the browser.
• Figure 3: ContextualEvent Generation: Users can steer the type of event generation within a single prompt (e.g. United States). (A) One context (World War II) can generate events focusing around that topic, while (B) another context (Revolution) can help constrain the generation to a completely separate group of events when the Events button is clicked.
• Figure 4: Generating Descriptions Flow: Users can (1) learn more details about a topic in the Search Box or event in the timeline by (2) using the Explain button. A detailed explanation is then (3) generated in the Side Panel for the given topic/event. To facilitate further exploration on the topic, the Questions button (4) can be clicked to generate a list of questions relating to the topic and context. Users can (5) select one of these questions to learn more about and a response will be (6) generated again on the Side Panel. (A) Source citations are available for users to verify the accuracy of the information.
• Figure 5: Legend Panel: All Event Types and their assigned color coding are listed here. Users can filter and efficiently navigate (b) to a select group of nodes corresponding to the label they select. Matching nodes of that type are highlighted and the view zooms in to focus on these events.