The Inter-American Development Bank (IDB) provides financial and technical support to improve lives in the Americas. Beyond funding loans and development projects across South America and the Caribbean, IDB generates thousands of pieces of original research, on-line courses, blog posts, briefs on countries and projects. All of these resources are available to internal staff, but they were difficult to find within four distinct search applications. Because of these information silos, even if a researcher found a relevant article they would have no idea that there was a project brief on the same topic just a few clicks away.
To increase the utilization of their proprietary datasets, IDB built a unification layer than can connect concepts across data silos. Using Natural Language Processing and Stardog’s built-in machine learning, the knowledge graph derives which content is the most similar to the search query (or the article the user is currently reading) based on context and content, not just tags and keywords. Given any piece of proprietary research or any set of search terms, the knowledge graph can determine the next best resource.
Through the Findability Project, the most relevant results are curated and delivered to each searcher no matter where the data lives within the organization. By embedding the context into the data through a comprehensive data model, IDB powers five distinct applications (including a chatbot) using the same knowledge graph. Bank employees can now find the information they need, when they need it, wherever they are.