Replication Data: Leveraging Researcher Domain Expertise to Annotate Concepts within Imbalanced Data

In this manuscript, we describe a method to utilize researcher domain expertise to annotate concepts efficiently and accurately within an imbalanced dataset. This folder contains two scripts that run two variations of the simulation referred to in our paper. Additionally, we included two separate datasets that were utilized in the simulations. For each, we shared the list of document embeddings used for classification, together with a corresponding CSV which holds the categorical labels for each embedding. We recommend first reading the "README" text file, before running the scripts.