Data-driven path filtering in ConceptNet

Abstract

In many applications, it is important to characterize the way in which two concepts are semantically related. Knowledge graphs such as ConceptNet provide a rich source of information for such characterizations by encoding relations between concepts as edges in a graph. When two concepts are not directly connected by an edge, their relationship can still be described in terms of the paths that connect them. Unfortunately, many of these paths are uninformative and noisy, meaning that the success of applications that use such path features crucially relies on their ability to select high-quality paths. In existing applications, this path selection process is based on relatively simple heuristics. In this thesis I instead propose to learn to predict path quality from crowdsourced human assessments. Since a generic task-independent notion of quality is concerned, human participants are asked to rank paths according to their subjective assessment of the paths' naturalness, without being given specific definitions or guidelines. Experiments show that a neural network model trained on these assessments is able to predict human judgments on unseen paths with near optimal performance. Most notably, the resulting path selection method is substantially better than the current heuristic approaches at identifying meaningful paths in various applications.