In the current era of knowledge explosion, it is becoming increasingly critical to help researchers quickly grasp the core ideas and methods used in the sea of published articles. As a first step toward the aim, this paper proposes a novel approach that simulates the cognitive process of how human beings read Earth science articles, and automatically identifies semantic entities from the articles. Among others, one major objective is to identify the datasets studied in articles. Oftentimes, however,researchersdonotexplicitlycitethedatasetsused.Thus, we propose a profile-matching method strengthened by a neural network-based method to identify implicitly cited dataset entities based on the context. Our experiments have demonstrated the effectiveness of our approaches.