Yale University introduces LUX, a free and innovative research tool designed to lead users down a captivating rabbit hole of infinite connections surrounding their subjects of interest. Serving as a central hub, LUX offers access to a vast collection of 17 million searchable objects from Yale’s museums, archives, and libraries. While traditional search engines typically provide links to external sites, LUX goes beyond mere hits by establishing relationships between the searched object and other relevant items in the collection, revealing obscure connections.

Working in a manner akin to a search engine, LUX uncovers a wealth of information beyond the objects themselves. For instance, when searching for a specific artwork, LUX surfaces works from the same artist, pieces created during the same era or in the same location, as well as related art, books, and more. In the past, users had to visit multiple venues or conduct separate Google searches to piece together these resources. LUX consolidates these diverse elements into a single platform, simplifying research endeavors.

At the core of LUX lies a knowledge graph, a sophisticated backend data model that organizes information from various sources into a network of relationships. Similar to the pin-up boards used by detectives to visualize connections between people, objects, places, and events, knowledge graphs facilitate comprehensive context and understanding. This approach gained prominence with Google in 2012, and Van Gogh World Wide operates on a similar data model. The art world is increasingly embracing this technique as digitization progresses.

Robert Sanderson, senior director for digital cultural heritage at Yale University, emphasizes that LUX offers robust context surrounding the searched object, delivering a comprehensive research experience. Tabs on the LUX page categorize searches into different sections, such as objects, works, people and groups, places, concepts, and events. Advanced search options and filters enable users to refine their queries further. Clicking through to a page reveals hyperlinks that unveil cross-connections, enabling users to explore related concepts, timelines, individuals, and more.

The development of LUX has been an ongoing project for the past five years. Yale University aims to simplify the process for other institutions to create their versions of LUX by open-sourcing the code. While the searching database remains proprietary, it can be licensed. Additionally, a smaller, similar database will be made available to smaller institutions with limited resources.

Notably, LUX relies on human intelligence rather than artificial intelligence. The team enlisted students to meticulously curate the metadata and add identifiers to datasets within the collections over a span of six years. Although experiments with ChatGPT were conducted, the reliance on human intelligence proved more reliable, as language models alone lacked the comprehensive knowledge and accuracy required.

While LUX is currently accessible to the public, it continues to evolve and improve. The team behind LUX is actively working on enhancing the tool and considering additional features. Users can provide valuable feedback through a prominent blue button on result pages, allowing them to report ethical concerns or inaccuracies in the data.

LUX stands as a testament to Yale University’s commitment to empowering research and enabling users to explore the boundless connections within their chosen subjects, revolutionizing the research experience for scholars and enthusiasts alike.

By Impact Lab