CHS: Small: Collaborative Research: RUI: Human-centered Semantic Relatedness

Semantic relatedness (SR) measures help computers replicate human assessments of the relatedness between two concepts. Decades of research and development have transformed SR measures into a critical component of a wide swath of intelligent technologies in areas ranging from information retrieval to human-computer interaction to spatial computing. Despite the importance and ubiquity of SR, researchers have only recently begun to examine it from a human-centered perspective. These human-centered studies have problematized key assumptions underlying the entire SR literature, e.g. that people from all cultural contexts agree on a single relatedness value between any two concepts. Our proposal addresses long-overdue open questions that will move the field towards a human-centered approach to semantic relatedness. To do so, we will collect datasets of relatedness judgments, mine patterns in Wikipedia content, perform statistical analyses, create and evaluate algorithms, develop software, and conduct large-scale user studies. We will first engage in four threads of work that redefine semantic relatedness to address human-centered concerns raised in the SR literature: (1) we will investigate the role of culture in SR and use these results to redefine SR to incorporate cultural context, (2) we will study SR among the low-notability concepts that are critical to end users but entirely ignored by the SR literature, (3) we will address the call for SR measures that explain their relatedness estimates to users, and (4) we will develop more robust human-centered SR evaluation procedures and support their adoption through easy-to-use software. Second, we will develop new conceptual representations for SR measures that accommodate differing cultural perspectives and create compact contextual SR models that empower applications with tractable human-centered SR algorithms. Finally, we will demonstrate the power of human-centered SR approaches through their application in improved recommender systems, enhanced Wikipedia reader experiences, and novel information discovery tools. Three themes span our proposed work: (1) we will disseminate our research advances through WikiBrain, a software library already used by hundreds of people around the world, (2) we will conduct live lab studies using active online systems we built and operate, and (3) we will work closely with large commercial and non-profit stakeholders (see attached letters from Thomson-Reuters and the Wikimedia Foundation).